{
  "schemaVersion": 1,
  "kind": "gezel-scorecard-evidence",
  "scorecardFile": "scorecard.json",
  "generatedAt": "2026-10-22T13:20:28.824Z",
  "campaign": {
    "status": "closed-at-user-request",
    "originalPlannedTrials": 903,
    "recordedTrials": 601,
    "unrunTrials": 302,
    "rawOutcomeCounts": {
      "recorded": 601,
      "passed": 476,
      "taskFailures": 119,
      "infrastructureExclusions": 6
    },
    "scorecardOutcomeCounts": {
      "recorded": 600,
      "passed": 476,
      "infrastructureExclusions": 6,
      "taskFailures": 118
    },
    "quarantinedObservations": [
      {
        "trialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "failureClass": "model",
        "reason": "Frozen comparison changed: fixtures: \"auto-answer.ts\", \"batch.ts\", \"bin/ab-work-routing.ts\", \"bin/image-bench.ts\", \"craftbooks/adapter.ts\", \"craftbooks/authoring/author-linear.ts\", \"craftbooks/authoring/edit-midtask.ts\", \"craftbooks/authoring/export-generalize.ts\", and 40 more; runtime: \"@bendyline/gezel-mcp/package.json\", \"@bendyline/gezel-service/dependency/@bendyline/docblocks-cli\", \"@bendyline/gezel/dist/browser.js\", \"@bendyline/gezel/dist/chunk-3Z6PBVZM.js\", \"@bendyline/gezel/dist/chunk-5BXY37IX.js\", \"@bendyline/gezel/dist/chunk-6ASB7JVJ.js\", \"@bendyline/gezel/dist/chunk-6RY2SMVG.js\", \"@bendyline/gezel/dist/chunk-6YKM4W6M.js\", and 28 more; source: \"evals/src/auto-answer.ts\", \"evals/src/batch.ts\", \"evals/src/bin/ab-work-routing.ts\", \"evals/src/bin/image-bench.ts\", \"evals/src/craftbooks/adapter.ts\", \"evals/src/craftbooks/authoring/author-linear.ts\", \"evals/src/craftbooks/authoring/edit-midtask.ts\", \"evals/src/craftbooks/authoring/export-generalize.ts\", and 40 more (see identity-drift-after-trial.json)",
        "explanation": "Retained without reclassification or replay; omitted only from the schema-v1 comparable dataset because comparison identity changed during this observation."
      }
    ],
    "unrunModels": [
      {
        "modelId": "nemotron3.5-lightning-30b-q4",
        "recordedTrials": 0,
        "plannedTrials": 129
      },
      {
        "modelId": "minicpm-v4.7-35b-a3b-q4",
        "recordedTrials": 0,
        "plannedTrials": 129
      }
    ],
    "rawScoresChanged": false,
    "newInferenceCalls": 0,
    "newJudgeCalls": 0,
    "environmentRestored": false,
    "monitorStatus": "PAUSED"
  },
  "runGroups": [
    {
      "runId": "2026-10-21-m5max-ornith-observed-subset",
      "scope": "subset",
      "modelIds": [
        "ornith1.5-35b-a3b-q4"
      ],
      "recordedTrials": 84,
      "sourceHash": "d68acf8def595c54deea3468ab9efe7ff2c6970be4e1d64e6fe034dbb82dd042",
      "fixtureHash": "ef50db65e73e3caf72db6f2e506cd0a397233fc4ab6eaaf0f6e2a52e360abdb3",
      "runtimeHash": "0329c67db45ad54224e7fc42574e8ea24dcf176bbf1ac5516d6d8dea813fbb57"
    },
    {
      "runId": "2026-10-16-m5max-local-seven-completed-arms",
      "scope": "suite",
      "modelIds": [
        "btl4-35b-q4",
        "lfm2.5-24b-a2b-q4",
        "qwen3.6-35b-a3b-q4",
        "qwen3.8-27b-q4"
      ],
      "recordedTrials": 516,
      "sourceHash": "d68acf8def595c54deea3468ab9efe7ff2c6970be4e1d64e6fe034dbb82dd042",
      "fixtureHash": "ef50db65e73e3caf72db6f2e506cd0a397233fc4ab6eaaf0f6e2a52e360abdb3",
      "runtimeHash": "0329c67db45ad54224e7fc42574e8ea24dcf176bbf1ac5516d6d8dea813fbb57"
    }
  ],
  "judging": {
    "judgeProvider": "codex-inline",
    "judgeModel": "current-session",
    "advisory": true,
    "artifactLimitBytes": 20480,
    "rubricInstructions": "Integer scores 0–10; tend toward 5 for delivering the brief, reserve 9–10 for polished work; specific justification at most 400 characters.",
    "coverage": {
      "recordedTrials": 601,
      "reviewedTrials": 601,
      "judged": 248,
      "noArtifact": 4,
      "invalidRubric": 14,
      "noDeclaredRubric": 335,
      "pending": 0
    },
    "crossRubricAggregation": false,
    "failedProducedArtifactsIncluded": true,
    "standardScorecardJudgeOmittedReason": "Schema v1 offers only a single model/suite judge mean. These task rubrics are not pooled. Use rubricAggregates and the per-trial judge entries instead.",
    "rubricDefinitions": [
      {
        "id": "core/tictactoe",
        "suiteId": "core",
        "scenarioId": "tictactoe",
        "originalPrompt": "Create a new project for this work (e.g. `Tic-Tac-Toe Game`) and build a tic-tac-toe game I can play in my browser IN THAT NEW PROJECT. Single HTML file at `workspace/index.html`, no build step. Two players take turns; show the winner when the game ends. Prefer nine static `<button class=\"cell\" data-cell=\"0\">` through `data-cell=\"8\"` elements in the HTML before the script, then wire them with `document.querySelectorAll(\".cell\")`. If you generate the board dynamically, create exactly 9 cells with a numeric loop such as `for (let i = 0; i < 9; i++)`; do not iterate an initially empty cells array to create the cells.",
        "description": "Meester creates a project, recruits a developer, and ships a working tic-tac-toe HTML game.",
        "artifactKind": "html",
        "artifactBasename": "index.html",
        "contextNote": null,
        "axes": [
          {
            "name": "visualQuality",
            "description": "Does it look like a real product, or a debug page?"
          },
          {
            "name": "functionalCompleteness",
            "description": "Does it implement the user's stated brief?"
          },
          {
            "name": "codeQuality",
            "description": "Would a developer reading the source be embarrassed?"
          },
          {
            "name": "polish",
            "description": "Are there surprising / delightful touches?"
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "core/petshop",
        "suiteId": "core",
        "scenarioId": "petshop",
        "originalPrompt": "Build me a single-page pet shop website with a custom logo. Create the shipping `index.html` in the website workspace first, with the logo wired in as <img src=\"assets/logo.png\" alt=\"Pet shop logo\"> from the first write. For the logo I want an AI-generated image (PNG file) — please use the image-generation tool to render it, save it in the website workspace as `assets/logo.png` using `generate_image({ saveAs: \"assets/logo.png\" })`, and reference it from an <img src=\"assets/logo.png\"> tag. Use one project/workspace for the whole website and logo; do not create a separate logo-only project. Do not wait to write the HTML until after image generation; the Developer/Builder owns `index.html`, while any image specialist only owns `assets/logo.png`. No build step, no inline SVG substitution.",
        "description": "Meester orchestrates a multi-tool job: developer builds a pet shop HTML site, image-generator renders a custom logo, and the page references it.",
        "artifactKind": "html",
        "artifactBasename": "index.html",
        "contextNote": null,
        "axes": [
          {
            "name": "visualQuality",
            "description": "Does it look like a real product, or a debug page?"
          },
          {
            "name": "functionalCompleteness",
            "description": "Does it implement the user's stated brief?"
          },
          {
            "name": "codeQuality",
            "description": "Would a developer reading the source be embarrassed?"
          },
          {
            "name": "polish",
            "description": "Are there surprising / delightful touches?"
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "core/tankcombat",
        "suiteId": "core",
        "scenarioId": "tankcombat",
        "originalPrompt": "Create a new project for this work (e.g. `Tank Combat Arcade`) and build a top-down tank combat arcade game there. Write the entire game as a single self-contained file at `workspace/index.html` IN THAT NEW PROJECT (HTML + inline `<style>` + inline `<script>`, no external assets, no build step). The player controls a tank with arrow keys (or WASD), fires projectiles with space, and fights at least one enemy tank. Track a score on screen. Make keyboard input observable in the browser: expose `window.gameState = gameState` or update a `document.body.dataset.inputTick` value inside key handlers so automated runtime checks can see input activity. The first concrete deliverable is the index.html file inside the new project — write it now. TANK_COMBAT_HTML_SHAPE: write one complete `index.html` document in a single workspace `write_file` call. Use exactly this source shape: `<!doctype html><html><head><style>...</style></head><body>...<canvas id=\"game\"></canvas>...<script>/* all game JavaScript here */</script></body></html>`. All JavaScript must be inside that one inline `<script>` block before `</body>`: declarations, functions, querySelector/addEventListener calls, key handlers, requestAnimationFrame game loop, projectiles, collision, score, and `window.gameState = gameState`. Do not place `const`, `let`, `var`, `function`, `querySelector`, `addEventListener`, or `requestAnimationFrame` after `</script>`, after `</body>`, or after `</html>`. If `write_file` is rejected for JavaScript outside a script tag, do not keep patching fragments; immediately re-emit the entire corrected HTML file from `<!doctype html>` through `</html>` with all JS inside the script block.",
        "description": "Meester creates a project, recruits a developer, and ships a playable top-down tank combat arcade game in a single HTML file.",
        "artifactKind": "html",
        "artifactBasename": "index.html",
        "contextNote": null,
        "axes": [
          {
            "name": "visualQuality",
            "description": "Does it look like a real product, or a debug page?"
          },
          {
            "name": "functionalCompleteness",
            "description": "Does it implement the user's stated brief?"
          },
          {
            "name": "codeQuality",
            "description": "Would a developer reading the source be embarrassed?"
          },
          {
            "name": "polish",
            "description": "Are there surprising / delightful touches?"
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/meeting-followup",
        "suiteId": "productivity",
        "scenarioId": "meeting-followup",
        "originalPrompt": "Noor is preparing the follow-up in the \"Orchid Launch Follow-up\" project. No Meester action is needed; just acknowledge this note.",
        "description": "Meeting-to-execution task: reconcile a noisy transcript with a stale agenda and current roster, then produce a decision brief plus an exact machine-readable action register with owners, dates, dependencies, unresolved questions, and risks preserved.",
        "artifactKind": "markdown",
        "artifactBasename": "meeting-brief.md",
        "contextNote": "The August 27 transcript is authoritative; old-agenda.md contains rejected proposals that must not become decisions.",
        "axes": [
          {
            "name": "decisionClarity",
            "description": "Separates decisions, open questions, actions, and risks cleanly."
          },
          {
            "name": "grounding",
            "description": "Owners, dates, scope, and dependencies match the transcript and roster."
          },
          {
            "name": "actionability",
            "description": "A reader can execute the follow-up without rereading the transcript."
          },
          {
            "name": "brevity",
            "description": "The brief is scannable and avoids transcript-style repetition."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "scenarioId": "craftbook-spreadsheet-model",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/craftbook-eval-brief.md\" })`, `read_file({ path: \"source/brief.md\" })`, `read_file({ path: \"source/records.csv\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverable with `write_file` at the exact path: `index.html` (not `workspace/<path>`).\nWrite the text artifact deliverable with `write_artifact` at the exact artifacts-drawer path: `scope.md` in this task's own artifacts working folder. Re-read it with `read_artifact`; do not create a workspace copy.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nUse the `spreadsheet-model` (Spreadsheet / Financial Model) craftbook for this self-contained local evaluation. Open `source/craftbook-eval-brief.md` first, then follow the craftbook's actual steps and gates through its terminal step. Complete this request: Build a working spreadsheet-style model (budget, forecast, unit economics, pricing, or scenario calculator) with clearly separated inputs, calculations, and outputs. Scopes the model's purpose, drivers, and assumptions FIRST so the formulas have a contract, then builds the model with live formulas and labeled assumption cells, then verifies the math by hand-checking key outputs and a sanity scenario. Use this for a financial model, budget calculator, ROI/forecast spreadsheet, unit-economics model, or any input-driven calculator.\n\nA gallery craftbook generated from an archetype spec. It runs\n`phase → (per-phase gate) → … → evaluate → (loop) → finish`. Each build\nphase that produces a checkable artifact is followed by a **runtime\ngate-checkpoint** — the runtime verifies the artifact and routes with no\nmodel turn, looping back to redo the phase on a miss. The final `evaluate`\nstep holds a static deliverable gate plus a reviewer QA pass. What it adds\nover the generic `build-loop`: a specialist role per phase, a\ndomain-correct ordering, and a concrete per-phase quality bar.\n\nDeliverables marked \"artifact\" land in the project's artifacts drawer (`write_artifact` / `read_artifact`), not the shipped workspace — review output is not product source.\n\nPhases:\n\n1. Scope the model (planner) — lock the purpose, drivers, assumptions, and outputs → gated on artifact `{{workPath}}/scope.md` (markdown-notes)\n2. Build the model (developer) — implement the model with live formulas and labeled cells → gated on `index.html` (html-page)\n\nThe gates never advance with an unmet criterion, and loop back to the\nowning phase to fix named gaps.\n Produce `scope.md` in this task's artifact working folder, `index.html`. Use only seeded project files and installed local simulators. Do not call a real external service, fetch live data, or claim an external side effect occurred; an explicit dry-run record is the correct result when a dependency is unavailable.",
        "description": "Spreadsheet / Financial Model workflow processability: Prove that a medium local model can parse and execute the Spreadsheet / Financial Model craftbook, satisfy its declared runtime gates, and reach a terminal step using only local fixtures.",
        "artifactKind": "markdown",
        "artifactBasename": "scope.md",
        "contextNote": "Advisory only. Deterministic pass/fail comes from the craftbook runtime gates, deliverable existence, attribution, and terminal progress.",
        "axes": [
          {
            "name": "workflow fidelity",
            "description": "The result reflects the named craftbook sequence and its declared acceptance gates rather than a freehand substitute."
          },
          {
            "name": "usefulness",
            "description": "The deliverable is coherent, actionable, and suitable for the craftbook purpose stated in the local brief."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "scenarioId": "docblocks-theme-roundtrip",
        "originalPrompt": "Wieke is aligning the quarterly review deck in the \"Brand Theme Alignment\" project. No Meester action is needed; just acknowledge this note.",
        "description": "DocBlocks theme round-trip: infer a theme from an existing brand document, apply it, re-inspect, compare revisions, and publish a report whose every cited value (theme id, two fonts, two hex colors, before/after page counts, the one unresolved style) came from a tool response. Grades tool provenance, a real PPTX container, and grounding against invented facts no model can guess.",
        "artifactKind": "markdown",
        "artifactBasename": "theme-report.md",
        "contextNote": "The DocBlocks tools returned exactly these values: theme id BRAND-THEME-7F2A; heading font Kelder Grotesk; body font Marne Text; accent #2E5E4E; background #F4F1EA; 11 pages before and 9 after; and one style that would not transfer, \"Callout Emphasis\". A deterministic gate already checks that each of these appears, so judge what it cannot: whether the report adds detail the tools never returned, states the unresolved style plainly, and ends somewhere useful. Do not mark a report down for lacking per-fact tool attribution — the brief does not ask for it.",
        "axes": [
          {
            "name": "grounding",
            "description": "No invented detail: every specific claim is one of the values listed above, or is plainly hedged. Numbers, fonts, or colors beyond that list are fabrication."
          },
          {
            "name": "candor",
            "description": "The unresolved style conflict is stated plainly rather than smoothed over."
          },
          {
            "name": "usefulness",
            "description": "The recommendation is a decision someone could act on this week."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "scenarioId": "craftbook-powerpoint-deck",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/brief.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverable with `write_file` at the exact path: `powerpoint/eval/deck.md` (not `workspace/<path>`).\nWrite the text artifact deliverables with `write_artifact` at the exact artifacts-drawer paths: `sources.md` in this task's own artifacts working folder, `outline.md` in this task's own artifacts working folder. Re-read it with `read_artifact`; do not create a workspace copy.\nProduce the real binary deliverable through the active craftbook production workflow: `deliverables/d-day.pptx`. Author and review the text source first, then use DocBlocks `convert_document`, `preview_document`, and `save_artifact`; finally use `copy_artifact_to_workspace` to copy the saved binary bytes to each exact workspace path. Never call `write_file` with prose, base64, or hand-built OOXML for these binary paths, and do not replace the craftbook with an ad-hoc Developer handoff.\nProduce the real binary artifact through the active craftbook workflow: `deck.pptx` in this task's own artifacts working folder. Author and review the source, then use DocBlocks `convert_document`, `preview_document`, and `save_artifact` so the exact binary path remains in the artifacts drawer. Do not copy it into the workspace and do not substitute prose, base64, or hand-built OOXML.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nCreate a PowerPoint from source/brief.md for operations leads. Use 6-8 slides, preserve the supplied facts exactly, and close with next actions. Do not create HTML.\n\nProduce the finished file through the DocBlocks tools on your roster: `list_roots`, then `convert_document` on the approved `powerpoint/eval/deck.md`, `preview_document` to confirm it rendered, save_artifact to retain the deck in the task artifacts folder, and copy_artifact_to_workspace to save it at deliverables/d-day.pptx. Inspect the saved file and finish by telling the user where it is. Markdown or hand-built OOXML written to that path is not a substitute for the real file.",
        "description": "PowerPoint source grounding, outline parity, and final saved file: Measure the complete PowerPoint workflow through its terminal save step: acquire and cite sources, match the approved outline, convert and inspect the PPTX, and retain both the artifact and requested workspace copy without a separate evaluation or summary handoff.",
        "artifactKind": "markdown",
        "artifactBasename": "deck.md",
        "contextNote": null,
        "axes": [
          {
            "name": "source-acquisition",
            "description": "The workflow read and cited the exact supplied source before outlining."
          },
          {
            "name": "outline-parity",
            "description": "Every numbered outline slide appears exactly once as a same-title H1 in powerpoint/eval/deck.md, in order."
          },
          {
            "name": "grounding",
            "description": "Every metric and action comes from source/brief.md and no extra metric was invented."
          },
          {
            "name": "convertibility",
            "description": "One H1 per slide and concise block content make the source ready for DocBlocks PPTX conversion."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/wikipedia-research-brief",
        "suiteId": "productivity",
        "scenarioId": "wikipedia-research-brief",
        "originalPrompt": "Elian is preparing the research brief in the \"Analytical Engine Research Brief\" project. No Meester action is needed; just acknowledge this note.",
        "description": "Hermetic tool-backed research: discover and call a local Wikipedia MCP, select the three relevant signed source cards out of four returned hits, synthesize them into a 700-1,500-word cited brief, preserve chronology, and calibrate a disputed historical claim. Live retrieval is blocked by configuration AND asserted against in the grader.",
        "artifactKind": "markdown",
        "artifactBasename": "research-brief.md",
        "contextNote": "All historical evidence comes from the three signed local Wikipedia source cards. The “first programmer” label must remain calibrated rather than asserted as uncontested fact.",
        "axes": [
          {
            "name": "synthesis",
            "description": "Builds an argument across sources rather than listing source summaries."
          },
          {
            "name": "grounding",
            "description": "Claims and dates stay faithful to the signed local source cards."
          },
          {
            "name": "citationQuality",
            "description": "Source ids appear near the claims they support and make provenance easy to audit."
          },
          {
            "name": "calibration",
            "description": "Separates documented facts, interpretation, and disputed historical shorthand."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/craftbook-week-plan",
        "suiteId": "productivity",
        "scenarioId": "craftbook-week-plan",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"followups.json\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverable with `write_file` at the exact path: `tasks/eval/report.md` (not `workspace/<path>`).\nUse the requested craftbook/template as guidance or by invoking it, but do not stop after creating a task. Execute the active craftbook step(s) now until the required output exists.\nmonday again — can you pull the week from the calender service and lay out a plan in tasks/eval/report.md? theres a probe script wired to the fake calendar, and my open followups are in the workspace too",
        "description": "Plan the Week smoke eval: Task-specific eval for the Plan the Week craftbook: the week's eight events come from a live fake calendar service, and tasks/eval/report.md shapes the days with the agenda-less meetings flagged for prep, the Tuesday conflict named, and focus blocks defended.",
        "artifactKind": "markdown",
        "artifactBasename": "report.md",
        "contextNote": null,
        "axes": [
          {
            "name": "coverage",
            "description": "Every event from the calendar week appears in the plan; nothing is silently dropped and nothing is invented."
          },
          {
            "name": "prioritization",
            "description": "Focus blocks land in real gaps, the biggest work gets the biggest gap, and follow-ups sit on days with room."
          },
          {
            "name": "conflict-handling",
            "description": "The overlapping pair is named with a concrete proposed fix the owner can accept in one decision."
          },
          {
            "name": "realism",
            "description": "The load picture is honest — heavy days are called heavy, and no day is over-packed to look tidy."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/craftbook-ab-test-readout",
        "suiteId": "productivity",
        "scenarioId": "craftbook-ab-test-readout",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/experiment-brief.md\" })`, `read_file({ path: \"data/checkout-test.csv\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverables with `write_file` at the exact paths: `data/stats.json`, `tasks/eval/readout.md` (not `workspace/<path>`).\nUse the requested craftbook/template as guidance or by invoking it, but do not stop after creating a task. Execute the active craftbook step(s) now until the required output exists.\nIn the `A/B Test Readout Eval` project, read `source/experiment-brief.md` and `data/checkout-test.csv`, then produce two files.\n\n1. `data/stats.json` — the computed figures, using EXACTLY this schema and these units:\n```json\n{\n  \"control\": { \"exposures\": 0, \"conversions\": 0, \"rate\": 0.0, \"refund_requests\": 0, \"refund_rate\": 0.0 },\n  \"variant\":  { \"exposures\": 0, \"conversions\": 0, \"rate\": 0.0, \"refund_requests\": 0, \"refund_rate\": 0.0 },\n  \"absolute_lift_pp\": 0.0,\n  \"relative_lift_pct\": 0.0,\n  \"p_value\": 0.0,\n  \"significant_at_95\": false,\n  \"guardrail_delta_pp\": 0.0,\n  \"guardrail_breached\": false,\n  \"decision\": \"ship\" | \"do_not_ship\" | \"inconclusive\"\n}\n```\n`rate` and `refund_rate` are FRACTIONS (0.12, not 12). `absolute_lift_pp` and `guardrail_delta_pp` are PERCENTAGE POINTS (1.25, not 0.0125). `relative_lift_pct` is a percentage (10.4). Use a two-proportion z-test for `p_value`. Round the two `_pp` fields and `relative_lift_pct` to 2 decimal places and the rate fields to 4, so a computed float does not serialize as 1.2500000000000009.\n\n2. `tasks/eval/readout.md` — a decision-ready experiment readout built from those figures and no new numbers. Lead with Ship / Do not ship / Inconclusive, report the conversion lift with confidence/significance language, check the refund-request guardrail, and tie the recommendation to the pre-registered decision rule. Apply the decision rule as written: every clause must hold to ship.",
        "description": "Decision-ready A/B test readout: Measure whether the ab-test-readout craftbook computes the experiment figures CORRECTLY (checked against a harness-owned arithmetic oracle, not prose keyword matching), and whether it applies a decision rule whose guardrail clause overrides a positive, statistically significant primary metric.",
        "artifactKind": "markdown",
        "artifactBasename": "readout.md",
        "contextNote": null,
        "axes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "applicability": "judgeSpecMismatch"
      },
      {
        "id": "productivity/craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "scenarioId": "craftbook-annotated-bibliography",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/source-cards.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverable with `write_file` at the exact path: `bibliography.md` (not `workspace/<path>`).\nUse the requested craftbook/template as guidance or by invoking it, but do not stop after creating a task. Execute the active craftbook step(s) now until the required output exists.\nIn the `Annotated Bibliography Eval` project, read `source/source-cards.md` and write `bibliography.md` for a research question about rural telehealth follow-up programs. Use APA-style entries and give every source a Summary, Critical evaluation, and Relevance annotation. Do not invent sources beyond the six seeded cards.",
        "description": "Annotated bibliography from seeded source cards: Measure whether the annotated-bibliography craftbook locks style/rubric and produces uniform summary, evaluation, and relevance annotations.",
        "artifactKind": "markdown",
        "artifactBasename": "bibliography.md",
        "contextNote": null,
        "axes": [
          {
            "name": "accuracy",
            "description": "Computed figures are correct against the seeded records."
          },
          {
            "name": "grounding",
            "description": "Every claim traces to the seeded data — nothing invented."
          },
          {
            "name": "structure",
            "description": "Tables and sections make the analysis checkable at a glance."
          },
          {
            "name": "actionability",
            "description": "Recommendations follow from the findings and are concretely doable."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "productivity/craftbook-research-to-document",
        "suiteId": "productivity",
        "scenarioId": "craftbook-research-to-document",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/brief.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverable with `write_file` at the exact path: `documents/eval/report.md` (not `workspace/<path>`).\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nRead `source/brief.md` and publish a Word document.\n\n1. Write the complete document source to `documents/eval/report.md`: a title, executive summary, compact metrics table, findings, and next actions. Do not create HTML or document XML by hand.\n2. Publish it with the pre-authorized DocBlocks tools. Call `list_roots` to find the read-enabled workspace root and the write-enabled artifacts root. Use `autoTemplates: true`; consult `get_authoring_context` only if exact authoring guidance is needed. Then `convert_document` to DOCX, `preview_document` to confirm it rendered, and `save_artifact` to persist it to `tasks/eval/report.docx` in the artifacts drawer.\n\nThe finished deliverable is the real DOCX at `tasks/eval/report.docx` in the artifacts drawer. Markdown, HTML, or hand-built OOXML written to that path is not a substitute.",
        "description": "Word Document DocBlocks artifact-surface eval: Verify the research-to-document craftbook carries a reviewed Markdown source all the way to a real editable DOCX in the ARTIFACTS drawer via the DocBlocks convert / preview / save workflow — the surface and tool family powerpoint-deck's workspace-copy route does not cover.",
        "artifactKind": "markdown",
        "artifactBasename": "report.md",
        "contextNote": null,
        "axes": [
          {
            "name": "grounding",
            "description": "Every metric and action comes from the seeded brief."
          },
          {
            "name": "structure",
            "description": "The source has a usable title, summary, sections, table, and next actions."
          },
          {
            "name": "convertibility",
            "description": "The Markdown is complete enough for direct DocBlocks DOCX conversion without HTML or XML."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "developer/craftbook-code-review",
        "suiteId": "developer",
        "scenarioId": "craftbook-code-review",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/craftbook-eval-brief.md\" })`, `read_file({ path: \"src/payment.js\" })`, `read_file({ path: \"docs/review-conventions.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text artifact deliverable with `write_artifact` at the exact artifacts-drawer path: `reviews/rev-eval-1/report.md`. Re-read it with `read_artifact`; do not create a workspace copy.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nUse the `code-review` (Code Review) craftbook for this self-contained local evaluation. Open `source/craftbook-eval-brief.md` first, then follow the craftbook's actual steps and gates through its terminal step. Complete this request: Review one staged change set and deliver a gated report.\n\nThe change set is snapshotted ahead of time into the project artifacts drawer at `reviews/<reviewId>/`: `manifest.json` (kind, base/head, changed files, and — for branch reviews — the commit list) plus `changes.diff` (the unified diff). Three steps:\n\n1. **Scope** — read the snapshot, note the review kind and the themes to check\n2. **Report** — walk the diff and write `reviews/<reviewId>/report.md` (Summary, severity-ranked Findings table with file:line citations, Verdict)\n3. **Done** — stamp the verdict to task notes\n\nThe reviewer never modifies project source — the only deliverable is the report artifact. Reports must ground every finding in the actual diff; the completion gate checks structure, citations, and verdict consistency. Produce `reviews/rev-eval-1/report.md`. Use only seeded project files and installed local simulators. Do not call a real external service, fetch live data, or claim an external side effect occurred; an explicit dry-run record is the correct result when a dependency is unavailable.",
        "description": "Code Review workflow processability: Prove that a medium local model can parse and execute the Code Review craftbook, satisfy its declared runtime gates, and reach a terminal step using only local fixtures.",
        "artifactKind": "markdown",
        "artifactBasename": "report.md",
        "contextNote": "Advisory only. Deterministic pass/fail comes from the craftbook runtime gates, deliverable existence, attribution, and terminal progress.",
        "axes": [
          {
            "name": "workflow fidelity",
            "description": "The result reflects the named craftbook sequence and its declared acceptance gates rather than a freehand substitute."
          },
          {
            "name": "usefulness",
            "description": "The deliverable is coherent, actionable, and suitable for the craftbook purpose stated in the local brief."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "developer/craftbook-api-contract-review",
        "suiteId": "developer",
        "scenarioId": "craftbook-api-contract-review",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"source/craftbook-eval-brief.md\" })`, `read_file({ path: \"openapi.yaml\" })`, `read_file({ path: \"src/routes.ts\" })`, `read_file({ path: \"src/session-routes.ts\" })`, `read_file({ path: \"docs/api-contract.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text artifact deliverables with `write_artifact` at the exact artifacts-drawer paths: `tasks/eval/api-review.md`, `tasks/eval/review.md`, `tasks/eval/scope.md`. Re-read it with `read_artifact`; do not create a workspace copy.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nUse the `api-contract-review` (API Contract Review) craftbook for this self-contained local evaluation. Open `source/craftbook-eval-brief.md` first, then follow the craftbook's actual steps and gates through its terminal step. Complete this request: Review an API contract or specification (OpenAPI/REST/GraphQL) for correctness, consistency, and developer ergonomics, then produce a findings report. First scopes the surface and locks an API design checklist (resource naming and HTTP semantics, status codes, pagination and filtering, error schema consistency, auth, versioning and backward compatibility, idempotency, and schema/type rigor), then reviews each endpoint and type against those rules and REST/GraphQL best practices, then writes a findings report rating each issue and giving the corrected contract snippet. Use this for an API design review, reviewing an OpenAPI/Swagger spec, a REST or GraphQL contract review, endpoint/schema critique, or checking an API for consistency and backward compatibility before it ships — it returns specific, spec-anchored fixes rather than 'make the API cleaner'.\n\nA gallery craftbook generated from an archetype spec. It runs\n`phase → (per-phase gate) → … → evaluate → (loop) → finish`. Each build\nphase that produces a checkable artifact is followed by a **runtime\ngate-checkpoint** — the runtime verifies the artifact and routes with no\nmodel turn, looping back to redo the phase on a miss. The final `evaluate`\nstep holds a static deliverable gate plus a reviewer QA pass. What it adds\nover the generic `build-loop`: a specialist role per phase, a\ndomain-correct ordering, and a concrete per-phase quality bar.\n\nDeliverables marked \"artifact\" land in the project's artifacts drawer (`write_artifact` / `read_artifact`), not the shipped workspace — review output is not product source.\n\nPhases:\n\n1. Scope the contract review (reviewer) — map the surface, lock an API design checklist → gated on artifact `{{workPath}}/scope.md` (markdown-notes)\n2. Review the endpoints (developer) — check each operation + type against the rules → gated on artifact `{{workPath}}/review.md` (markdown-notes)\n3. Write the contract report (reviewer) — findings with corrected snippets + compat verdict → gated on artifact `{{workPath}}/api-review.md` (markdown-report)\n\nThe gates never advance with an unmet criterion, and loop back to the\nowning phase to fix named gaps.\n Produce `tasks/eval/api-review.md`, `tasks/eval/review.md`, `tasks/eval/scope.md`. Use only seeded project files and installed local simulators. Do not call a real external service, fetch live data, or claim an external side effect occurred; an explicit dry-run record is the correct result when a dependency is unavailable.",
        "description": "API Contract Review workflow processability: Prove that a medium local model can parse and execute the API Contract Review craftbook, satisfy its declared runtime gates, and reach a terminal step using only local fixtures.",
        "artifactKind": "markdown",
        "artifactBasename": "api-review.md",
        "contextNote": "Advisory only. Deterministic pass/fail comes from the craftbook runtime gates, deliverable existence, attribution, and terminal progress.",
        "axes": [
          {
            "name": "workflow fidelity",
            "description": "The result reflects the named craftbook sequence and its declared acceptance gates rather than a freehand substitute."
          },
          {
            "name": "usefulness",
            "description": "The deliverable is coherent, actionable, and suitable for the craftbook purpose stated in the local brief."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "developer/craftbook-deep-security-review",
        "suiteId": "developer",
        "scenarioId": "craftbook-deep-security-review",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"src/admin.ts\" })`, `read_file({ path: \"src/profile.ts\" })`, `read_file({ path: \"src/reports.ts\" })`, `read_file({ path: \"docs/security.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverables with `write_file` at the exact paths: `security-review/REPORT.md`, `security-review/findings.json` (not `workspace/<path>`).\nUse the requested craftbook/template as guidance or by invoking it, but do not stop after creating a task. Execute the active craftbook step(s) now until the required output exists.\nUse the Deep Security Review craftbook for a whole-codebase security sweep. The fixture is offline and self-contained. Write security-review/REPORT.md and security-review/findings.json; do not modify source.",
        "description": "Deep Security Review systemic eval: Measure whether the review confirms seeded source-to-sink vulnerabilities, identifies systemic root causes, and emits valid machine-readable findings.",
        "artifactKind": "markdown",
        "artifactBasename": "REPORT.md",
        "contextNote": null,
        "axes": [
          {
            "name": "verification",
            "description": "Raw leads are confirmed as real source-to-sink or authorization paths."
          },
          {
            "name": "systemic analysis",
            "description": "Findings are grouped by root cause and blast radius."
          },
          {
            "name": "remediation",
            "description": "The report and JSON contain concrete, severity-consistent fixes."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "developer/craftbook-codemod-sweep",
        "suiteId": "developer",
        "scenarioId": "craftbook-codemod-sweep",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"package.json\" })`, `read_file({ path: \"src/api.js\" })`, `read_file({ path: \"src/index.js\" })`, `read_file({ path: \"src/profile.js\" })`, `read_file({ path: \"tests/profile.test.mjs\" })`, `read_file({ path: \"CHANGELOG.md\" })`, `read_file({ path: \"source/rename-brief.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text artifact deliverables with `write_artifact` at the exact artifacts-drawer paths: `sites.md` in this task's own artifacts working folder, `change-notes.md` in this task's own artifacts working folder, `verification.md` in this task's own artifacts working folder, `review.md` in this task's own artifacts working folder. Re-read it with `read_artifact`; do not create a workspace copy.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nUse the Codemod Sweep craftbook for this: rename fetchUserData to loadUser across the whole repo - the export, the re-export, every import and call site, and log strings that mention the old name. CHANGELOG.md is history and keeps the old name. The brief is in source/rename-brief.md. No behavior change.",
        "description": "Repo-wide rename sweep with reconciled counts: The craftbook drives a complete mechanical rename of fetchUserData to loadUser across a small repo whose sites include a re-export and a template-literal log string, with a changelog legitimately excluded as history - verified mechanically by a residual scan, behavior probes on every site, and a green suite.",
        "artifactKind": "markdown",
        "artifactBasename": "verification.md",
        "contextNote": null,
        "axes": [
          {
            "name": "Enumeration completeness",
            "description": "Every site was found by real searches before any edit - the definition, the re-export, the imports and call sites, and the log string - with the changelog excluded for a stated reason rather than missed."
          },
          {
            "name": "Invariant discipline",
            "description": "The change stayed mechanical: nothing changed beyond the declared rename, the excluded files are untouched, and behavior is identical."
          },
          {
            "name": "Reconciled verification",
            "description": "The residual re-scan is quoted with its result, the found/changed/excluded counts add up exactly, and the suite run backing the green claim is real."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "developer/craftbook-refactor-module",
        "suiteId": "developer",
        "scenarioId": "craftbook-refactor-module",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"package.json\" })`, `read_file({ path: \"src/format-name.js\" })`, `read_file({ path: \"src/report.js\" })`, `read_file({ path: \"tests/format.test.mjs\" })`, `read_file({ path: \"tests/report.test.mjs\" })`, `read_file({ path: \"source/refactor-brief.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text artifact deliverables with `write_artifact` at the exact artifacts-drawer paths: `baseline.md` in this task's own artifacts working folder, `plan.md` in this task's own artifacts working folder, `change-notes.md` in this task's own artifacts working folder, `verification.md` in this task's own artifacts working folder, `review.md` in this task's own artifacts working folder. Re-read it with `read_artifact`; do not create a workspace copy.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nUse the Behavior-Preserving Refactor craftbook for this: the same name-normalization logic is copy-pasted in src/format-name.js and src/report.js, and the copies have drifted once already. The brief is in source/refactor-brief.md. Extract the shared logic into a new src/normalize-name.js used by both files, without changing behavior.",
        "description": "Behavior-preserving extraction of duplicated logic: The craftbook drives a disciplined refactor of copy-pasted name-normalization logic: a green baseline recorded before anything moves, reversible stages with the suite green throughout, a re-measured duplication target, and an enforced review — verified mechanically via a reference matrix across both call sites, a duplication count across src/, and a green-suite run.",
        "artifactKind": "markdown",
        "artifactBasename": "change-notes.md",
        "contextNote": null,
        "axes": [
          {
            "name": "Guardrail-first discipline",
            "description": "The suite was genuinely run and recorded green before any restructuring, characterization coverage was considered honestly, and the baseline counted the duplication with a stated method."
          },
          {
            "name": "Reversible-stage execution",
            "description": "The restructure happened in small planned stages at the real sites, with the suite kept green between stages and deviations recorded honestly rather than papered over."
          },
          {
            "name": "Genuine deduplication",
            "description": "One shared implementation remains, both call sites use it, public exports survived unchanged, and no behavior change was smuggled in under cover of the refactor."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "developer/craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "scenarioId": "craftbook-bug-fix-tdd",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"package.json\" })`, `read_file({ path: \"src/pricing.js\" })`, `read_file({ path: \"tests/pricing.test.mjs\" })`, `read_file({ path: \"source/bug-report.md\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text artifact deliverables with `write_artifact` at the exact artifacts-drawer paths: `repro.md` in this task's own artifacts working folder, `diagnosis.md` in this task's own artifacts working folder, `fix-notes.md` in this task's own artifacts working folder, `validation.md` in this task's own artifacts working folder, `review.md` in this task's own artifacts working folder. Re-read it with `read_artifact`; do not create a workspace copy.\nUse the requested craftbook/template by invoking it when needed, but keep the work task-native: create and complete the authoring flow for the draft task graph. Do not activate the draft and do not build the planned workspace artifact.\nUse the Reproduce-Then-Fix a Bug craftbook for this: customers report that buying exactly 10 of an item does not get the bulk discount, but the pricing page promises the discount at 10 or more. The report is in source/bug-report.md. Reproduce it with a failing test first, then fix it properly.",
        "description": "Bug fix with red-then-green proof: The craftbook drives a disciplined fix of a seeded boundary bug: a failing reproduction test written into the real test layout, the smallest change at the real defect site, a green suite afterwards, and an enforced review — all verified mechanically, including a mutant run proving the regression guard bites.",
        "artifactKind": "markdown",
        "artifactBasename": "fix-notes.md",
        "contextNote": null,
        "axes": [
          {
            "name": "Repro-first discipline",
            "description": "A failing test was genuinely written and run before the fix, in the project's real test layout, and the red evidence quotes a real run rather than a claim."
          },
          {
            "name": "Fix minimality",
            "description": "The change is the smallest correct edit at the real defect site; no special-casing of the reported input, no unrelated churn."
          },
          {
            "name": "Honest verification",
            "description": "Validation re-ran the original reproduction, distinguishes verified results from anything unverified, and the review graded evidence rather than summaries."
          }
        ],
        "applicability": "applicable"
      },
      {
        "id": "complex-work/craftbook-invoice-run",
        "suiteId": "complex-work",
        "scenarioId": "craftbook-invoice-run",
        "originalPrompt": "[craftbook eval harness]\nThis project is self-contained; all referenced source paths are workspace-root-relative.\nBefore writing any deliverable, read every seeded input with workspace file tools: `read_file({ path: \"clients.json\" })`, `read_file({ path: \"ledger.json\" })`. Do not use `read_artifact`, `read_document`, or library tools for these workspace files.\nWrite the text workspace deliverable with `write_file` at the exact path: `report.md` (not `workspace/<path>`).\nUse the requested craftbook/template as guidance or by invoking it, but do not stop after creating a task. Execute the active craftbook step(s) now until the required output exists.\nend of month — can you run the invoicing? three clients have billable work logged, last invoice number was 2026-041, dont break the sequnce",
        "description": "Monthly Invoice Run — ledger to numbered, recorded invoices: Measure whether the invoice-run craftbook turns a seeded client roster and work ledger into one print-ready invoice per billable client with unbroken numbering, ledger-true amounts, and a run summary — numeric integrity is the class bar.",
        "artifactKind": "markdown",
        "artifactBasename": "report.md",
        "contextNote": null,
        "axes": [
          {
            "name": "numeric integrity",
            "description": "Every amount traces to a tracked ledger entry; nothing estimated or invented."
          },
          {
            "name": "numbering continuity",
            "description": "Invoice numbers continue 2026-041 with no skip and no repeat."
          },
          {
            "name": "professionalism",
            "description": "The invoices look worth paying — clear line items, terms-correct due dates, print-ready pages."
          },
          {
            "name": "completeness",
            "description": "One invoice per billable client, the idle client skipped on purpose, and the books updated."
          },
          {
            "name": "print-polish",
            "description": "Each invoice is styled for paper — an @page rule, honest margins, a professional face."
          }
        ],
        "applicability": "applicable"
      }
    ],
    "rubricAggregates": [
      {
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "axes": [
          "numeric integrity",
          "numbering continuity",
          "professionalism",
          "completeness",
          "print-polish"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "numeric integrity": 2.3333333333333335,
              "numbering continuity": 8,
              "professionalism": 4.333333333333333,
              "completeness": 4.333333333333333,
              "print-polish": 6
            },
            "withinRubricMean": 5,
            "trialIds": [
              "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut",
              "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz",
              "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "numeric integrity": 7.333333333333333,
              "numbering continuity": 8,
              "professionalism": 5.333333333333333,
              "completeness": 4.333333333333333,
              "print-polish": 8
            },
            "withinRubricMean": 6.6,
            "trialIds": [
              "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9",
              "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9",
              "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "numeric integrity": 8.666666666666666,
              "numbering continuity": 9,
              "professionalism": 5.666666666666667,
              "completeness": 4,
              "print-polish": 8
            },
            "withinRubricMean": 7.066666666666666,
            "trialIds": [
              "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i",
              "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs",
              "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "numeric integrity": 8.333333333333334,
              "numbering continuity": 9,
              "professionalism": 2,
              "completeness": 4,
              "print-polish": 5
            },
            "withinRubricMean": 5.666666666666667,
            "trialIds": [
              "craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk",
              "craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x",
              "craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "numeric integrity": 9,
              "numbering continuity": 9,
              "professionalism": 4,
              "completeness": 4,
              "print-polish": 6.5
            },
            "withinRubricMean": 6.5,
            "trialIds": [
              "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p",
              "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v"
            ]
          }
        }
      },
      {
        "suite": "core",
        "scenario": "petshop",
        "axes": [
          "visualQuality",
          "functionalCompleteness",
          "codeQuality",
          "polish"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 5.333333333333333,
              "functionalCompleteness": 5.333333333333333,
              "codeQuality": 5.666666666666667,
              "polish": 3.3333333333333335
            },
            "withinRubricMean": 4.916666666666667,
            "trialIds": [
              "petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki",
              "petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef",
              "petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 7.333333333333333,
              "functionalCompleteness": 6,
              "codeQuality": 6.333333333333333,
              "polish": 5.333333333333333
            },
            "withinRubricMean": 6.25,
            "trialIds": [
              "petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s",
              "petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik",
              "petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 6.666666666666667,
              "codeQuality": 6.666666666666667,
              "polish": 5
            },
            "withinRubricMean": 6.083333333333333,
            "trialIds": [
              "petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl",
              "petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4",
              "petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 4.666666666666667,
              "functionalCompleteness": 6,
              "codeQuality": 5,
              "polish": 3.6666666666666665
            },
            "withinRubricMean": 4.833333333333333,
            "trialIds": [
              "petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv",
              "petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh",
              "petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "visualQuality": 7,
              "functionalCompleteness": 7,
              "codeQuality": 7,
              "polish": 5.5
            },
            "withinRubricMean": 6.625,
            "trialIds": [
              "petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4",
              "petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex"
            ]
          }
        }
      },
      {
        "suite": "core",
        "scenario": "tankcombat",
        "axes": [
          "visualQuality",
          "functionalCompleteness",
          "codeQuality",
          "polish"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 4.333333333333333,
              "functionalCompleteness": 1,
              "codeQuality": 2.3333333333333335,
              "polish": 2
            },
            "withinRubricMean": 2.4166666666666665,
            "trialIds": [
              "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr",
              "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao",
              "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 5.333333333333333,
              "codeQuality": 5.333333333333333,
              "polish": 5
            },
            "withinRubricMean": 5.416666666666667,
            "trialIds": [
              "tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao",
              "tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh",
              "tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 8,
              "codeQuality": 6.666666666666667,
              "polish": 5.666666666666667
            },
            "withinRubricMean": 6.583333333333333,
            "trialIds": [
              "tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a",
              "tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c",
              "tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 5,
              "functionalCompleteness": 7.333333333333333,
              "codeQuality": 5.666666666666667,
              "polish": 4.666666666666667
            },
            "withinRubricMean": 5.666666666666667,
            "trialIds": [
              "tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm",
              "tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp",
              "tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "visualQuality": 5.5,
              "functionalCompleteness": 2,
              "codeQuality": 2.5,
              "polish": 3.5
            },
            "withinRubricMean": 3.375,
            "trialIds": [
              "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb",
              "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn"
            ]
          }
        }
      },
      {
        "suite": "core",
        "scenario": "tictactoe",
        "axes": [
          "visualQuality",
          "functionalCompleteness",
          "codeQuality",
          "polish"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 4,
              "functionalCompleteness": 6.666666666666667,
              "codeQuality": 6,
              "polish": 3.3333333333333335
            },
            "withinRubricMean": 5,
            "trialIds": [
              "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1",
              "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o",
              "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 8,
              "codeQuality": 7.666666666666667,
              "polish": 5.666666666666667
            },
            "withinRubricMean": 6.833333333333333,
            "trialIds": [
              "tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke",
              "tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm",
              "tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 8,
              "codeQuality": 7,
              "polish": 5
            },
            "withinRubricMean": 6.5,
            "trialIds": [
              "tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs",
              "tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj",
              "tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 8,
              "codeQuality": 7,
              "polish": 5.333333333333333
            },
            "withinRubricMean": 6.583333333333333,
            "trialIds": [
              "tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm",
              "tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt",
              "tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "visualQuality": 6,
              "functionalCompleteness": 9,
              "codeQuality": 7,
              "polish": 6
            },
            "withinRubricMean": 7,
            "trialIds": [
              "tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w",
              "tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f"
            ]
          }
        }
      },
      {
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "axes": [
          "workflow fidelity",
          "usefulness"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 5,
              "usefulness": 2.3333333333333335
            },
            "withinRubricMean": 3.6666666666666665,
            "trialIds": [
              "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf",
              "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh",
              "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 5,
              "usefulness": 4.666666666666667
            },
            "withinRubricMean": 4.833333333333333,
            "trialIds": [
              "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj",
              "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf",
              "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 7.666666666666667,
              "usefulness": 5.666666666666667
            },
            "withinRubricMean": 6.666666666666667,
            "trialIds": [
              "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5",
              "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe",
              "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 7.666666666666667,
              "usefulness": 4
            },
            "withinRubricMean": 5.833333333333333,
            "trialIds": [
              "craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs",
              "craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl",
              "craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "workflow fidelity": 9,
              "usefulness": 6
            },
            "withinRubricMean": 7.5,
            "trialIds": [
              "craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0",
              "craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z"
            ]
          }
        }
      },
      {
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "axes": [
          "Repro-first discipline",
          "Fix minimality",
          "Honest verification"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "Repro-first discipline": 5.333333333333333,
              "Fix minimality": 8,
              "Honest verification": 5.666666666666667
            },
            "withinRubricMean": 6.333333333333333,
            "trialIds": [
              "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b",
              "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy",
              "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "Repro-first discipline": 7.333333333333333,
              "Fix minimality": 8,
              "Honest verification": 7.333333333333333
            },
            "withinRubricMean": 7.555555555555555,
            "trialIds": [
              "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1",
              "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u",
              "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "Repro-first discipline": 9,
              "Fix minimality": 9,
              "Honest verification": 8
            },
            "withinRubricMean": 8.666666666666666,
            "trialIds": [
              "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x",
              "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv",
              "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "Repro-first discipline": 9,
              "Fix minimality": 7,
              "Honest verification": 5.333333333333333
            },
            "withinRubricMean": 7.111111111111111,
            "trialIds": [
              "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme",
              "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m",
              "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "Repro-first discipline": 9,
              "Fix minimality": 8,
              "Honest verification": 6.5
            },
            "withinRubricMean": 7.833333333333333,
            "trialIds": [
              "craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk",
              "craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx"
            ]
          }
        }
      },
      {
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "axes": [
          "workflow fidelity",
          "usefulness"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 1,
            "axisMeans": {
              "workflow fidelity": 5,
              "usefulness": 3
            },
            "withinRubricMean": 4,
            "trialIds": [
              "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 5.333333333333333,
              "usefulness": 6
            },
            "withinRubricMean": 5.666666666666667,
            "trialIds": [
              "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh",
              "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib",
              "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 8,
              "usefulness": 7.333333333333333
            },
            "withinRubricMean": 7.666666666666667,
            "trialIds": [
              "craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214",
              "craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667",
              "craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 6.666666666666667,
              "usefulness": 6.666666666666667
            },
            "withinRubricMean": 6.666666666666667,
            "trialIds": [
              "craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7",
              "craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb",
              "craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "workflow fidelity": 7.5,
              "usefulness": 9
            },
            "withinRubricMean": 8.25,
            "trialIds": [
              "craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z",
              "craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt"
            ]
          }
        }
      },
      {
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "axes": [
          "Enumeration completeness",
          "Invariant discipline",
          "Reconciled verification"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "Enumeration completeness": 6,
              "Invariant discipline": 7.333333333333333,
              "Reconciled verification": 5.666666666666667
            },
            "withinRubricMean": 6.333333333333333,
            "trialIds": [
              "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v",
              "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq",
              "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "Enumeration completeness": 7,
              "Invariant discipline": 8,
              "Reconciled verification": 7
            },
            "withinRubricMean": 7.333333333333333,
            "trialIds": [
              "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd",
              "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur",
              "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "Enumeration completeness": 9,
              "Invariant discipline": 9,
              "Reconciled verification": 9
            },
            "withinRubricMean": 9,
            "trialIds": [
              "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124",
              "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc",
              "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "Enumeration completeness": 7.333333333333333,
              "Invariant discipline": 9,
              "Reconciled verification": 5
            },
            "withinRubricMean": 7.111111111111111,
            "trialIds": [
              "craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb",
              "craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195",
              "craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "Enumeration completeness": 9,
              "Invariant discipline": 9,
              "Reconciled verification": 9
            },
            "withinRubricMean": 9,
            "trialIds": [
              "craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba",
              "craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht"
            ]
          }
        }
      },
      {
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "axes": [
          "verification",
          "systemic analysis",
          "remediation"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "verification": 7,
              "systemic analysis": 5.666666666666667,
              "remediation": 5.333333333333333
            },
            "withinRubricMean": 6,
            "trialIds": [
              "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv",
              "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n",
              "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "verification": 7.333333333333333,
              "systemic analysis": 6,
              "remediation": 6.333333333333333
            },
            "withinRubricMean": 6.555555555555555,
            "trialIds": [
              "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz",
              "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4",
              "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "verification": 7.666666666666667,
              "systemic analysis": 6.666666666666667,
              "remediation": 7
            },
            "withinRubricMean": 7.111111111111111,
            "trialIds": [
              "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1",
              "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn",
              "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "verification": 8,
              "systemic analysis": 6,
              "remediation": 7.666666666666667
            },
            "withinRubricMean": 7.222222222222222,
            "trialIds": [
              "craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula",
              "craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55",
              "craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "verification": 8,
              "systemic analysis": 4,
              "remediation": 8
            },
            "withinRubricMean": 6.666666666666667,
            "trialIds": [
              "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw",
              "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k"
            ]
          }
        }
      },
      {
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "axes": [
          "Guardrail-first discipline",
          "Reversible-stage execution",
          "Genuine deduplication"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "Guardrail-first discipline": 4.333333333333333,
              "Reversible-stage execution": 3.6666666666666665,
              "Genuine deduplication": 3.6666666666666665
            },
            "withinRubricMean": 3.888888888888889,
            "trialIds": [
              "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c",
              "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult",
              "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "Guardrail-first discipline": 1,
              "Reversible-stage execution": 1,
              "Genuine deduplication": 0
            },
            "withinRubricMean": 0.6666666666666666,
            "trialIds": [
              "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w",
              "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "Guardrail-first discipline": 9,
              "Reversible-stage execution": 9,
              "Genuine deduplication": 9
            },
            "withinRubricMean": 9,
            "trialIds": [
              "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw",
              "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek",
              "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9"
            ]
          },
          "btl4-35b-q4": {
            "n": 2,
            "axisMeans": {
              "Guardrail-first discipline": 3.5,
              "Reversible-stage execution": 5.5,
              "Genuine deduplication": 5.5
            },
            "withinRubricMean": 4.833333333333333,
            "trialIds": [
              "craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji",
              "craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "Guardrail-first discipline": 8.5,
              "Reversible-stage execution": 5,
              "Genuine deduplication": 9
            },
            "withinRubricMean": 7.5,
            "trialIds": [
              "craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab",
              "craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "axes": [
          "accuracy",
          "grounding",
          "structure",
          "actionability"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "accuracy": 5.666666666666667,
              "grounding": 4,
              "structure": 6,
              "actionability": 5.333333333333333
            },
            "withinRubricMean": 5.25,
            "trialIds": [
              "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r",
              "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2",
              "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "accuracy": 6,
              "grounding": 4,
              "structure": 7,
              "actionability": 5.666666666666667
            },
            "withinRubricMean": 5.666666666666667,
            "trialIds": [
              "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w",
              "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e",
              "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "accuracy": 7.333333333333333,
              "grounding": 4.666666666666667,
              "structure": 8,
              "actionability": 6.333333333333333
            },
            "withinRubricMean": 6.583333333333333,
            "trialIds": [
              "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d",
              "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1",
              "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "accuracy": 9,
              "grounding": 3,
              "structure": 9,
              "actionability": 6
            },
            "withinRubricMean": 6.75,
            "trialIds": [
              "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj",
              "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m",
              "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "accuracy": 9,
              "grounding": 4,
              "structure": 7.5,
              "actionability": 6.5
            },
            "withinRubricMean": 6.75,
            "trialIds": [
              "craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o",
              "craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "axes": [
          "source-acquisition",
          "outline-parity",
          "grounding",
          "convertibility"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "source-acquisition": 6.333333333333333,
              "outline-parity": 5.333333333333333,
              "grounding": 7.666666666666667,
              "convertibility": 5.333333333333333
            },
            "withinRubricMean": 6.166666666666667,
            "trialIds": [
              "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh",
              "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x",
              "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "source-acquisition": 7,
              "outline-parity": 8,
              "grounding": 5.333333333333333,
              "convertibility": 8
            },
            "withinRubricMean": 7.083333333333333,
            "trialIds": [
              "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii",
              "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6",
              "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "source-acquisition": 8.666666666666666,
              "outline-parity": 8.666666666666666,
              "grounding": 7,
              "convertibility": 7.666666666666667
            },
            "withinRubricMean": 8,
            "trialIds": [
              "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z",
              "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem",
              "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "source-acquisition": 9,
              "outline-parity": 9,
              "grounding": 6.666666666666667,
              "convertibility": 8
            },
            "withinRubricMean": 8.166666666666666,
            "trialIds": [
              "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3",
              "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz",
              "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "source-acquisition": 9,
              "outline-parity": 9,
              "grounding": 7,
              "convertibility": 9
            },
            "withinRubricMean": 8.5,
            "trialIds": [
              "craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r",
              "craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "axes": [
          "grounding",
          "structure",
          "convertibility"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 7.333333333333333,
              "structure": 7,
              "convertibility": 8
            },
            "withinRubricMean": 7.444444444444445,
            "trialIds": [
              "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh",
              "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl",
              "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 5.333333333333333,
              "structure": 7,
              "convertibility": 8
            },
            "withinRubricMean": 6.777777777777778,
            "trialIds": [
              "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65",
              "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1",
              "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 5.666666666666667,
              "structure": 8,
              "convertibility": 8
            },
            "withinRubricMean": 7.222222222222222,
            "trialIds": [
              "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7",
              "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi",
              "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 4,
              "structure": 8,
              "convertibility": 8
            },
            "withinRubricMean": 6.666666666666667,
            "trialIds": [
              "craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk",
              "craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm",
              "craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "grounding": 3.5,
              "structure": 8.5,
              "convertibility": 8
            },
            "withinRubricMean": 6.666666666666667,
            "trialIds": [
              "craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1",
              "craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "axes": [
          "workflow fidelity",
          "usefulness"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 5.333333333333333,
              "usefulness": 3.6666666666666665
            },
            "withinRubricMean": 4.5,
            "trialIds": [
              "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n",
              "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i",
              "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 5,
              "usefulness": 3
            },
            "withinRubricMean": 4,
            "trialIds": [
              "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn",
              "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78",
              "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 8,
              "usefulness": 6.666666666666667
            },
            "withinRubricMean": 7.333333333333333,
            "trialIds": [
              "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio",
              "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh",
              "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "workflow fidelity": 6.666666666666667,
              "usefulness": 3.3333333333333335
            },
            "withinRubricMean": 5,
            "trialIds": [
              "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k",
              "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w",
              "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "workflow fidelity": 7.5,
              "usefulness": 4
            },
            "withinRubricMean": 5.75,
            "trialIds": [
              "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz",
              "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "axes": [
          "coverage",
          "prioritization",
          "conflict-handling",
          "realism"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "coverage": 0,
              "prioritization": 1,
              "conflict-handling": 0.3333333333333333,
              "realism": 1
            },
            "withinRubricMean": 0.5833333333333334,
            "trialIds": [
              "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg",
              "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j",
              "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "coverage": 3.3333333333333335,
              "prioritization": 2,
              "conflict-handling": 2,
              "realism": 2
            },
            "withinRubricMean": 2.3333333333333335,
            "trialIds": [
              "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv",
              "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn",
              "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "coverage": 1,
              "prioritization": 1.3333333333333333,
              "conflict-handling": 0,
              "realism": 1
            },
            "withinRubricMean": 0.8333333333333334,
            "trialIds": [
              "craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj",
              "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v",
              "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "coverage": 3,
              "prioritization": 3.3333333333333335,
              "conflict-handling": 1.6666666666666667,
              "realism": 2.3333333333333335
            },
            "withinRubricMean": 2.5833333333333335,
            "trialIds": [
              "craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8",
              "craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv",
              "craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "coverage": 0.5,
              "prioritization": 2,
              "conflict-handling": 0.5,
              "realism": 2
            },
            "withinRubricMean": 1.25,
            "trialIds": [
              "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z",
              "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "axes": [
          "grounding",
          "candor",
          "usefulness"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 4.666666666666667,
              "candor": 6.666666666666667,
              "usefulness": 4.666666666666667
            },
            "withinRubricMean": 5.333333333333333,
            "trialIds": [
              "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g",
              "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v",
              "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 2.6666666666666665,
              "candor": 6.666666666666667,
              "usefulness": 6
            },
            "withinRubricMean": 5.111111111111111,
            "trialIds": [
              "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg",
              "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb",
              "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 6.666666666666667,
              "candor": 8,
              "usefulness": 7.333333333333333
            },
            "withinRubricMean": 7.333333333333333,
            "trialIds": [
              "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a",
              "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89",
              "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "grounding": 7.333333333333333,
              "candor": 9,
              "usefulness": 7.333333333333333
            },
            "withinRubricMean": 7.888888888888889,
            "trialIds": [
              "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro",
              "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa",
              "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "grounding": 8,
              "candor": 9,
              "usefulness": 8
            },
            "withinRubricMean": 8.333333333333334,
            "trialIds": [
              "docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9",
              "docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "meeting-followup",
        "axes": [
          "decisionClarity",
          "grounding",
          "actionability",
          "brevity"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "decisionClarity": 7,
              "grounding": 6,
              "actionability": 7,
              "brevity": 7.333333333333333
            },
            "withinRubricMean": 6.833333333333333,
            "trialIds": [
              "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut",
              "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz",
              "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "decisionClarity": 7,
              "grounding": 7,
              "actionability": 7.666666666666667,
              "brevity": 7
            },
            "withinRubricMean": 7.166666666666667,
            "trialIds": [
              "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss",
              "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo",
              "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "decisionClarity": 8,
              "grounding": 8.333333333333334,
              "actionability": 8.666666666666666,
              "brevity": 8
            },
            "withinRubricMean": 8.25,
            "trialIds": [
              "meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd",
              "meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz",
              "meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "decisionClarity": 9,
              "grounding": 9,
              "actionability": 9,
              "brevity": 9
            },
            "withinRubricMean": 9,
            "trialIds": [
              "meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq",
              "meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t",
              "meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "decisionClarity": 9,
              "grounding": 8,
              "actionability": 8,
              "brevity": 9
            },
            "withinRubricMean": 8.5,
            "trialIds": [
              "meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y",
              "meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9"
            ]
          }
        }
      },
      {
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "axes": [
          "synthesis",
          "grounding",
          "citationQuality",
          "calibration"
        ],
        "models": {
          "lfm2.5-24b-a2b-q4": {
            "n": 3,
            "axisMeans": {
              "synthesis": 4,
              "grounding": 2.3333333333333335,
              "citationQuality": 4,
              "calibration": 4
            },
            "withinRubricMean": 3.5833333333333335,
            "trialIds": [
              "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d",
              "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y",
              "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f"
            ]
          },
          "qwen3.6-35b-a3b-q4": {
            "n": 3,
            "axisMeans": {
              "synthesis": 4.333333333333333,
              "grounding": 3.3333333333333335,
              "citationQuality": 5.333333333333333,
              "calibration": 5.666666666666667
            },
            "withinRubricMean": 4.666666666666667,
            "trialIds": [
              "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby",
              "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q",
              "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh"
            ]
          },
          "qwen3.8-27b-q4": {
            "n": 3,
            "axisMeans": {
              "synthesis": 5.666666666666667,
              "grounding": 6.333333333333333,
              "citationQuality": 7.666666666666667,
              "calibration": 7.666666666666667
            },
            "withinRubricMean": 6.833333333333333,
            "trialIds": [
              "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o",
              "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip",
              "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe"
            ]
          },
          "btl4-35b-q4": {
            "n": 3,
            "axisMeans": {
              "synthesis": 4,
              "grounding": 4,
              "citationQuality": 6.666666666666667,
              "calibration": 4.666666666666667
            },
            "withinRubricMean": 4.833333333333333,
            "trialIds": [
              "wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex",
              "wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe",
              "wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt"
            ]
          },
          "ornith1.5-35b-a3b-q4": {
            "n": 2,
            "axisMeans": {
              "synthesis": 4.5,
              "grounding": 4.5,
              "citationQuality": 6,
              "calibration": 5.5
            },
            "withinRubricMean": 5.125,
            "trialIds": [
              "wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo",
              "wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg"
            ]
          }
        }
      }
    ]
  },
  "performance": {
    "standardScorecardMethod": "Arithmetic means of paired recorded admitted preflight rates; samples counts probes, not trials. The same model probe set is shared across suite rows.",
    "standardScorecardConfidence": "All nine probes generated fewer than 256 decode tokens each; preliminary admission-probe measurements, not sustained throughput.",
    "trialAggregateMethod": "Median, min, max and observed-row count of each trial’s engine-derived rates. Wall durations include startup, model copying and tool work. No whole-trial token/sec fallback is used.",
    "models": {
      "lfm2.5-24b-a2b-q4": {
        "label": "LFM2.5 (24B-A2B, Q4)",
        "preflight": {
          "aggregate": {
            "prefillTokensPerSec": 4711,
            "decodeTokensPerSec": 182.6,
            "samples": 2
          },
          "probes": [
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/preflight/preflight-lfm2-5-24b-a2b-q4-2026-10-16T04-34-25-418Z-oc39/preflight-report.json",
              "sourceSha256": "965391a69acfdc424caae3d62ffe2a615f61fa9c7d9e3ff22ba46e5968418ada",
              "createdAt": "2026-10-16T04:34:42.564Z",
              "trialId": "preflight-lfm2-5-24b-a2b-q4-2026-10-16T04-34-25-418Z-oc39",
              "prefillTokensPerSec": 4715.07,
              "decodeTokensPerSec": 183.85,
              "decodeTokens": 158,
              "caveats": [
                "158 decode tokens < 256; compare in-trial engine timings instead"
              ]
            },
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/preflight/preflight-lfm2-5-24b-a2b-q4-2026-10-17T05-03-19-676Z-tiim/preflight-report.json",
              "sourceSha256": "ac71559b5d4f90af28563524fdad3ea499d78e9440b1d8820cd78afb4dadaa3b",
              "createdAt": "2026-10-17T05:03:31.846Z",
              "trialId": "preflight-lfm2-5-24b-a2b-q4-2026-10-17T05-03-19-676Z-tiim",
              "prefillTokensPerSec": 4706.18,
              "decodeTokensPerSec": 181.4,
              "decodeTokens": 172,
              "caveats": [
                "172 decode tokens < 256; compare in-trial engine timings instead"
              ]
            }
          ],
          "confidence": "short-admission-probes"
        },
        "allRecorded": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 128,
          "wallDurationMs": {
            "n": 128,
            "median": 368345,
            "min": 22043,
            "max": 5406715
          },
          "successfulWallDurationMs": {
            "n": 77,
            "median": 186117,
            "min": 22043,
            "max": 1404854
          },
          "decodeTokensPerSec": {
            "n": 128,
            "median": 104.69,
            "min": 72.61,
            "max": 170.47
          },
          "prefillTokensPerSec": {
            "n": 128,
            "median": 2298.725,
            "min": 1445.72,
            "max": 4769.38
          },
          "peakRssMb": {
            "n": 128,
            "median": 18541,
            "min": 16651,
            "max": 27301
          },
          "toolCalls": {
            "n": 128,
            "median": 33.5,
            "min": 1,
            "max": 488
          }
        },
        "comparisonObservations": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 128,
          "wallDurationMs": {
            "n": 128,
            "median": 368345,
            "min": 22043,
            "max": 5406715
          },
          "successfulWallDurationMs": {
            "n": 77,
            "median": 186117,
            "min": 22043,
            "max": 1404854
          },
          "decodeTokensPerSec": {
            "n": 128,
            "median": 104.69,
            "min": 72.61,
            "max": 170.47
          },
          "prefillTokensPerSec": {
            "n": 128,
            "median": 2298.725,
            "min": 1445.72,
            "max": 4769.38
          },
          "peakRssMb": {
            "n": 128,
            "median": 18541,
            "min": 16651,
            "max": 27301
          },
          "toolCalls": {
            "n": 128,
            "median": 33.5,
            "min": 1,
            "max": 488
          }
        },
        "bySuite": {
          "core": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 32,
            "wallDurationMs": {
              "n": 32,
              "median": 188923.5,
              "min": 37350,
              "max": 956717
            },
            "successfulWallDurationMs": {
              "n": 24,
              "median": 130464,
              "min": 37350,
              "max": 956717
            },
            "decodeTokensPerSec": {
              "n": 32,
              "median": 108.48,
              "min": 85.71,
              "max": 170.47
            },
            "prefillTokensPerSec": {
              "n": 32,
              "median": 2410.925,
              "min": 1643.91,
              "max": 4769.38
            },
            "peakRssMb": {
              "n": 32,
              "median": 17977,
              "min": 16651,
              "max": 25658
            },
            "toolCalls": {
              "n": 32,
              "median": 11.5,
              "min": 1,
              "max": 66
            }
          },
          "productivity": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 39,
            "wallDurationMs": {
              "n": 39,
              "median": 282216,
              "min": 26968,
              "max": 5406715
            },
            "successfulWallDurationMs": {
              "n": 28,
              "median": 177280.5,
              "min": 26968,
              "max": 719674
            },
            "decodeTokensPerSec": {
              "n": 39,
              "median": 104.14,
              "min": 78.31,
              "max": 138.83
            },
            "prefillTokensPerSec": {
              "n": 39,
              "median": 2219.1,
              "min": 1446.73,
              "max": 3223.12
            },
            "peakRssMb": {
              "n": 39,
              "median": 18520,
              "min": 16747,
              "max": 23519
            },
            "toolCalls": {
              "n": 39,
              "median": 26,
              "min": 3,
              "max": 488
            }
          },
          "developer": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 30,
            "wallDurationMs": {
              "n": 30,
              "median": 661303.5,
              "min": 22043,
              "max": 4581963
            },
            "successfulWallDurationMs": {
              "n": 18,
              "median": 318808.5,
              "min": 22043,
              "max": 996632
            },
            "decodeTokensPerSec": {
              "n": 30,
              "median": 107.935,
              "min": 72.61,
              "max": 144.13
            },
            "prefillTokensPerSec": {
              "n": 30,
              "median": 2392.05,
              "min": 1573.31,
              "max": 3572.51
            },
            "peakRssMb": {
              "n": 30,
              "median": 19568,
              "min": 16824,
              "max": 27301
            },
            "toolCalls": {
              "n": 30,
              "median": 42.5,
              "min": 4,
              "max": 408
            }
          },
          "complex-work": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 27,
            "wallDurationMs": {
              "n": 27,
              "median": 995108,
              "min": 30637,
              "max": 3898569
            },
            "successfulWallDurationMs": {
              "n": 7,
              "median": 324207,
              "min": 30637,
              "max": 1404854
            },
            "decodeTokensPerSec": {
              "n": 27,
              "median": 96.6,
              "min": 73.44,
              "max": 123.78
            },
            "prefillTokensPerSec": {
              "n": 27,
              "median": 1976.91,
              "min": 1445.72,
              "max": 2789.64
            },
            "peakRssMb": {
              "n": 27,
              "median": 18257,
              "min": 16816,
              "max": 23633
            },
            "toolCalls": {
              "n": 27,
              "median": 57,
              "min": 3,
              "max": 245
            }
          }
        }
      },
      "qwen3.6-35b-a3b-q4": {
        "label": "Qwen 3.6 (35B-A3B, Q4)",
        "preflight": {
          "aggregate": {
            "prefillTokensPerSec": 1441,
            "decodeTokensPerSec": 133.9,
            "samples": 2
          },
          "probes": [
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/preflight/preflight-qwen3-6-35b-a3b-q4-2026-10-17T08-23-07-242Z-67fj/preflight-report.json",
              "sourceSha256": "b29ae28647555260095d8a1b10eba69fbc116ce9ca643e9f6ec6b8365b4cc75c",
              "createdAt": "2026-10-17T08:23:24.460Z",
              "trialId": "preflight-qwen3-6-35b-a3b-q4-2026-10-17T08-23-07-242Z-67fj",
              "prefillTokensPerSec": 1387.2,
              "decodeTokensPerSec": 138.19,
              "decodeTokens": 94,
              "caveats": [
                "94 decode tokens < 256; compare in-trial engine timings instead"
              ]
            },
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/preflight/preflight-qwen3-6-35b-a3b-q4-2026-10-18T08-43-09-123Z-woji/preflight-report.json",
              "sourceSha256": "06e3ca70d1bc3989cec195c9048e2cd1524a38a5901751a91455b99146ab8f24",
              "createdAt": "2026-10-18T08:43:21.322Z",
              "trialId": "preflight-qwen3-6-35b-a3b-q4-2026-10-18T08-43-09-123Z-woji",
              "prefillTokensPerSec": 1494.97,
              "decodeTokensPerSec": 129.7,
              "decodeTokens": 96,
              "caveats": [
                "96 decode tokens < 256; compare in-trial engine timings instead"
              ]
            }
          ],
          "confidence": "short-admission-probes"
        },
        "allRecorded": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 127,
          "wallDurationMs": {
            "n": 127,
            "median": 143413,
            "min": 22738,
            "max": 5195219
          },
          "successfulWallDurationMs": {
            "n": 111,
            "median": 122460,
            "min": 22738,
            "max": 5195219
          },
          "decodeTokensPerSec": {
            "n": 127,
            "median": 94.37,
            "min": 56.15,
            "max": 159.33
          },
          "prefillTokensPerSec": {
            "n": 127,
            "median": 1506.71,
            "min": 648.61,
            "max": 2735.96
          },
          "peakRssMb": {
            "n": 127,
            "median": 34097,
            "min": 29156,
            "max": 45655
          },
          "toolCalls": {
            "n": 127,
            "median": 19,
            "min": 2,
            "max": 336
          }
        },
        "comparisonObservations": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 127,
          "wallDurationMs": {
            "n": 127,
            "median": 143413,
            "min": 22738,
            "max": 5195219
          },
          "successfulWallDurationMs": {
            "n": 111,
            "median": 122460,
            "min": 22738,
            "max": 5195219
          },
          "decodeTokensPerSec": {
            "n": 127,
            "median": 94.37,
            "min": 56.15,
            "max": 159.33
          },
          "prefillTokensPerSec": {
            "n": 127,
            "median": 1506.71,
            "min": 648.61,
            "max": 2735.96
          },
          "peakRssMb": {
            "n": 127,
            "median": 34097,
            "min": 29156,
            "max": 45655
          },
          "toolCalls": {
            "n": 127,
            "median": 19,
            "min": 2,
            "max": 336
          }
        },
        "bySuite": {
          "core": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 33,
            "wallDurationMs": {
              "n": 33,
              "median": 57165,
              "min": 22738,
              "max": 308631
            },
            "successfulWallDurationMs": {
              "n": 31,
              "median": 55012,
              "min": 22738,
              "max": 238101
            },
            "decodeTokensPerSec": {
              "n": 33,
              "median": 115.13,
              "min": 78.14,
              "max": 159.33
            },
            "prefillTokensPerSec": {
              "n": 33,
              "median": 1738.48,
              "min": 1312.22,
              "max": 2735.96
            },
            "peakRssMb": {
              "n": 33,
              "median": 30277,
              "min": 29156,
              "max": 39615
            },
            "toolCalls": {
              "n": 33,
              "median": 5,
              "min": 2,
              "max": 52
            }
          },
          "productivity": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 39,
            "wallDurationMs": {
              "n": 39,
              "median": 143413,
              "min": 27004,
              "max": 5195219
            },
            "successfulWallDurationMs": {
              "n": 35,
              "median": 132704,
              "min": 27004,
              "max": 5195219
            },
            "decodeTokensPerSec": {
              "n": 39,
              "median": 94.87,
              "min": 73.35,
              "max": 131.96
            },
            "prefillTokensPerSec": {
              "n": 39,
              "median": 1549.04,
              "min": 1242.23,
              "max": 2340.6
            },
            "peakRssMb": {
              "n": 39,
              "median": 33475,
              "min": 29291,
              "max": 45388
            },
            "toolCalls": {
              "n": 39,
              "median": 18,
              "min": 5,
              "max": 336
            }
          },
          "developer": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 30,
            "wallDurationMs": {
              "n": 30,
              "median": 242051,
              "min": 27371,
              "max": 1799462
            },
            "successfulWallDurationMs": {
              "n": 26,
              "median": 153019.5,
              "min": 27371,
              "max": 841958
            },
            "decodeTokensPerSec": {
              "n": 30,
              "median": 93.49000000000001,
              "min": 56.15,
              "max": 116.8
            },
            "prefillTokensPerSec": {
              "n": 30,
              "median": 1465.54,
              "min": 648.61,
              "max": 1936.35
            },
            "peakRssMb": {
              "n": 30,
              "median": 34771,
              "min": 29305,
              "max": 41369
            },
            "toolCalls": {
              "n": 30,
              "median": 23.5,
              "min": 2,
              "max": 142
            }
          },
          "complex-work": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 25,
            "wallDurationMs": {
              "n": 25,
              "median": 808540,
              "min": 25059,
              "max": 3300770
            },
            "successfulWallDurationMs": {
              "n": 19,
              "median": 524110,
              "min": 25059,
              "max": 3300770
            },
            "decodeTokensPerSec": {
              "n": 25,
              "median": 75.82,
              "min": 62.31,
              "max": 118.26
            },
            "prefillTokensPerSec": {
              "n": 25,
              "median": 1261.28,
              "min": 847.45,
              "max": 2393.5
            },
            "peakRssMb": {
              "n": 25,
              "median": 38651,
              "min": 29439,
              "max": 45655
            },
            "toolCalls": {
              "n": 25,
              "median": 60,
              "min": 2,
              "max": 130
            }
          }
        }
      },
      "qwen3.8-27b-q4": {
        "label": "Qwen 3.8 (27B, Q4)",
        "preflight": {
          "aggregate": {
            "prefillTokensPerSec": 501,
            "decodeTokensPerSec": 36.7,
            "samples": 2
          },
          "probes": [
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/preflight/preflight-qwen3-8-27b-q4-2026-10-18T11-04-38-925Z-chph/preflight-report.json",
              "sourceSha256": "1b7ea229fd18d4bd448adcc42f3493cd7ed5f164b4973d6677fe8b402625ce09",
              "createdAt": "2026-10-18T11:04:57.117Z",
              "trialId": "preflight-qwen3-8-27b-q4-2026-10-18T11-04-38-925Z-chph",
              "prefillTokensPerSec": 626.61,
              "decodeTokensPerSec": 41.96,
              "decodeTokens": 104,
              "caveats": [
                "104 decode tokens < 256; compare in-trial engine timings instead"
              ]
            },
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/preflight/preflight-qwen3-8-27b-q4-2026-10-19T11-33-09-990Z-7cpf/preflight-report.json",
              "sourceSha256": "e1fdcb0c1444a0c8d1831818bf7c7aa89b835ff36f34f3ec2884c42207d4ca13",
              "createdAt": "2026-10-19T11:33:33.173Z",
              "trialId": "preflight-qwen3-8-27b-q4-2026-10-19T11-33-09-990Z-7cpf",
              "prefillTokensPerSec": 375.39,
              "decodeTokensPerSec": 31.48,
              "decodeTokens": 74,
              "caveats": [
                "74 decode tokens < 256; compare in-trial engine timings instead"
              ]
            }
          ],
          "confidence": "short-admission-probes"
        },
        "allRecorded": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 129,
          "wallDurationMs": {
            "n": 129,
            "median": 497444,
            "min": 42652,
            "max": 6622225
          },
          "successfulWallDurationMs": {
            "n": 116,
            "median": 394942,
            "min": 42652,
            "max": 6622225
          },
          "decodeTokensPerSec": {
            "n": 129,
            "median": 22.93,
            "min": 12,
            "max": 35.66
          },
          "prefillTokensPerSec": {
            "n": 129,
            "median": 390.27,
            "min": 206.17,
            "max": 530.06
          },
          "peakRssMb": {
            "n": 129,
            "median": 46124,
            "min": 36812,
            "max": 54102
          },
          "toolCalls": {
            "n": 129,
            "median": 18,
            "min": 1,
            "max": 142
          }
        },
        "comparisonObservations": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 129,
          "wallDurationMs": {
            "n": 129,
            "median": 497444,
            "min": 42652,
            "max": 6622225
          },
          "successfulWallDurationMs": {
            "n": 116,
            "median": 394942,
            "min": 42652,
            "max": 6622225
          },
          "decodeTokensPerSec": {
            "n": 129,
            "median": 22.93,
            "min": 12,
            "max": 35.66
          },
          "prefillTokensPerSec": {
            "n": 129,
            "median": 390.27,
            "min": 206.17,
            "max": 530.06
          },
          "peakRssMb": {
            "n": 129,
            "median": 46124,
            "min": 36812,
            "max": 54102
          },
          "toolCalls": {
            "n": 129,
            "median": 18,
            "min": 1,
            "max": 142
          }
        },
        "bySuite": {
          "core": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 33,
            "wallDurationMs": {
              "n": 33,
              "median": 168189,
              "min": 42652,
              "max": 1303224
            },
            "successfulWallDurationMs": {
              "n": 33,
              "median": 168189,
              "min": 42652,
              "max": 1303224
            },
            "decodeTokensPerSec": {
              "n": 33,
              "median": 25.96,
              "min": 17.66,
              "max": 33.3
            },
            "prefillTokensPerSec": {
              "n": 33,
              "median": 402.48,
              "min": 360.5,
              "max": 530.06
            },
            "peakRssMb": {
              "n": 33,
              "median": 41395,
              "min": 36814,
              "max": 50553
            },
            "toolCalls": {
              "n": 33,
              "median": 6,
              "min": 1,
              "max": 27
            }
          },
          "productivity": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 39,
            "wallDurationMs": {
              "n": 39,
              "median": 506547,
              "min": 62489,
              "max": 5407153
            },
            "successfulWallDurationMs": {
              "n": 37,
              "median": 497444,
              "min": 62489,
              "max": 2665244
            },
            "decodeTokensPerSec": {
              "n": 39,
              "median": 23.54,
              "min": 16.95,
              "max": 35.66
            },
            "prefillTokensPerSec": {
              "n": 39,
              "median": 393.54,
              "min": 360.05,
              "max": 453
            },
            "peakRssMb": {
              "n": 39,
              "median": 44772,
              "min": 36812,
              "max": 51616
            },
            "toolCalls": {
              "n": 39,
              "median": 17,
              "min": 1,
              "max": 142
            }
          },
          "developer": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 30,
            "wallDurationMs": {
              "n": 30,
              "median": 502172.5,
              "min": 78097,
              "max": 4678018
            },
            "successfulWallDurationMs": {
              "n": 27,
              "median": 566711,
              "min": 78097,
              "max": 4678018
            },
            "decodeTokensPerSec": {
              "n": 30,
              "median": 22.295,
              "min": 17.91,
              "max": 30.55
            },
            "prefillTokensPerSec": {
              "n": 30,
              "median": 393.33000000000004,
              "min": 206.17,
              "max": 427.54
            },
            "peakRssMb": {
              "n": 30,
              "median": 46495.5,
              "min": 37427,
              "max": 53668
            },
            "toolCalls": {
              "n": 30,
              "median": 22,
              "min": 2,
              "max": 80
            }
          },
          "complex-work": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 27,
            "wallDurationMs": {
              "n": 27,
              "median": 2024310,
              "min": 66096,
              "max": 6622225
            },
            "successfulWallDurationMs": {
              "n": 19,
              "median": 1973780,
              "min": 66096,
              "max": 6622225
            },
            "decodeTokensPerSec": {
              "n": 27,
              "median": 18.5,
              "min": 12,
              "max": 27.53
            },
            "prefillTokensPerSec": {
              "n": 27,
              "median": 333.56,
              "min": 231.27,
              "max": 390.9
            },
            "peakRssMb": {
              "n": 27,
              "median": 48108,
              "min": 37512,
              "max": 54102
            },
            "toolCalls": {
              "n": 27,
              "median": 44,
              "min": 2,
              "max": 81
            }
          }
        }
      },
      "btl4-35b-q4": {
        "label": "BTL-4 (35B-A3B, Q4)",
        "preflight": {
          "aggregate": {
            "prefillTokensPerSec": 2828,
            "decodeTokensPerSec": 119.4,
            "samples": 2
          },
          "probes": [
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/preflight/preflight-btl4-35b-q4-2026-10-20T00-48-49-039Z-5xna/preflight-report.json",
              "sourceSha256": "88884372ead89ceb0f956c750bd4f662b74f0316f8eb16cd2e886af884919357",
              "createdAt": "2026-10-20T00:52:55.512Z",
              "trialId": "preflight-btl4-35b-q4-2026-10-20T00-48-49-039Z-5xna",
              "prefillTokensPerSec": 2807.67,
              "decodeTokensPerSec": 119.41,
              "decodeTokens": 118,
              "caveats": [
                "118 decode tokens < 256; compare in-trial engine timings instead"
              ]
            },
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/preflight/preflight-btl4-35b-q4-2026-10-21T00-57-14-888Z-m29i/preflight-report.json",
              "sourceSha256": "de6ba3de81981869138b871bd4823c12c1f1600ee3ad9bee6d764f8cb783de22",
              "createdAt": "2026-10-21T01:01:20.535Z",
              "trialId": "preflight-btl4-35b-q4-2026-10-21T00-57-14-888Z-m29i",
              "prefillTokensPerSec": 2848.77,
              "decodeTokensPerSec": 119.32,
              "decodeTokens": 252,
              "caveats": [
                "252 decode tokens < 256; compare in-trial engine timings instead"
              ]
            }
          ],
          "confidence": "short-admission-probes"
        },
        "allRecorded": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 128,
          "wallDurationMs": {
            "n": 128,
            "median": 366634.5,
            "min": 253528,
            "max": 6398111
          },
          "successfulWallDurationMs": {
            "n": 98,
            "median": 346708,
            "min": 253528,
            "max": 6398111
          },
          "decodeTokensPerSec": {
            "n": 128,
            "median": 98.43,
            "min": 35.95,
            "max": 118.07
          },
          "prefillTokensPerSec": {
            "n": 128,
            "median": 2183.23,
            "min": 647.12,
            "max": 2951.1
          },
          "peakRssMb": {
            "n": 128,
            "median": 30615.5,
            "min": 26519,
            "max": 41753
          },
          "toolCalls": {
            "n": 128,
            "median": 19,
            "min": 2,
            "max": 899
          }
        },
        "comparisonObservations": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 128,
          "wallDurationMs": {
            "n": 128,
            "median": 366634.5,
            "min": 253528,
            "max": 6398111
          },
          "successfulWallDurationMs": {
            "n": 98,
            "median": 346708,
            "min": 253528,
            "max": 6398111
          },
          "decodeTokensPerSec": {
            "n": 128,
            "median": 98.43,
            "min": 35.95,
            "max": 118.07
          },
          "prefillTokensPerSec": {
            "n": 128,
            "median": 2183.23,
            "min": 647.12,
            "max": 2951.1
          },
          "peakRssMb": {
            "n": 128,
            "median": 30615.5,
            "min": 26519,
            "max": 41753
          },
          "toolCalls": {
            "n": 128,
            "median": 19,
            "min": 2,
            "max": 899
          }
        },
        "bySuite": {
          "core": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 33,
            "wallDurationMs": {
              "n": 33,
              "median": 311351,
              "min": 257373,
              "max": 503085
            },
            "successfulWallDurationMs": {
              "n": 28,
              "median": 287530.5,
              "min": 257373,
              "max": 503085
            },
            "decodeTokensPerSec": {
              "n": 33,
              "median": 103.45,
              "min": 54.64,
              "max": 115.37
            },
            "prefillTokensPerSec": {
              "n": 33,
              "median": 2449.67,
              "min": 1115.43,
              "max": 2951.1
            },
            "peakRssMb": {
              "n": 33,
              "median": 28769,
              "min": 27068,
              "max": 37576
            },
            "toolCalls": {
              "n": 33,
              "median": 8,
              "min": 2,
              "max": 45
            }
          },
          "productivity": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 39,
            "wallDurationMs": {
              "n": 39,
              "median": 376543,
              "min": 255682,
              "max": 5639168
            },
            "successfulWallDurationMs": {
              "n": 31,
              "median": 376543,
              "min": 255682,
              "max": 5639168
            },
            "decodeTokensPerSec": {
              "n": 39,
              "median": 99.27,
              "min": 51.64,
              "max": 118.07
            },
            "prefillTokensPerSec": {
              "n": 39,
              "median": 2281.58,
              "min": 1188.25,
              "max": 2924.03
            },
            "peakRssMb": {
              "n": 39,
              "median": 30710,
              "min": 26519,
              "max": 41753
            },
            "toolCalls": {
              "n": 39,
              "median": 19,
              "min": 3,
              "max": 899
            }
          },
          "developer": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 30,
            "wallDurationMs": {
              "n": 30,
              "median": 390928,
              "min": 255460,
              "max": 2427510
            },
            "successfulWallDurationMs": {
              "n": 25,
              "median": 317065,
              "min": 255460,
              "max": 1321208
            },
            "decodeTokensPerSec": {
              "n": 30,
              "median": 93.215,
              "min": 35.95,
              "max": 115.63
            },
            "prefillTokensPerSec": {
              "n": 30,
              "median": 2051.84,
              "min": 647.12,
              "max": 2850.95
            },
            "peakRssMb": {
              "n": 30,
              "median": 29953,
              "min": 27117,
              "max": 40125
            },
            "toolCalls": {
              "n": 30,
              "median": 22,
              "min": 2,
              "max": 206
            }
          },
          "complex-work": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 26,
            "wallDurationMs": {
              "n": 26,
              "median": 1490733,
              "min": 253528,
              "max": 6398111
            },
            "successfulWallDurationMs": {
              "n": 14,
              "median": 739213.5,
              "min": 253528,
              "max": 6398111
            },
            "decodeTokensPerSec": {
              "n": 26,
              "median": 67.03,
              "min": 48.2,
              "max": 109.37
            },
            "prefillTokensPerSec": {
              "n": 26,
              "median": 1300.79,
              "min": 815.61,
              "max": 2744.83
            },
            "peakRssMb": {
              "n": 26,
              "median": 34272.5,
              "min": 27308,
              "max": 38411
            },
            "toolCalls": {
              "n": 26,
              "median": 62.5,
              "min": 2,
              "max": 270
            }
          }
        }
      },
      "ornith1.5-35b-a3b-q4": {
        "label": "Ornith 1.5 (35B-A3B, Q4)",
        "preflight": {
          "aggregate": {
            "prefillTokensPerSec": 2619,
            "decodeTokensPerSec": 190.6,
            "samples": 1
          },
          "probes": [
            {
              "source": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/preflight/preflight-ornith1-5-35b-a3b-q4-2026-10-21T09-49-24-537Z-jnst/preflight-report.json",
              "sourceSha256": "816eee077dd14d16eb856964e37211d758325da2ef99aac69632bd144f6c0f25",
              "createdAt": "2026-10-21T09:53:34.650Z",
              "trialId": "preflight-ornith1-5-35b-a3b-q4-2026-10-21T09-49-24-537Z-jnst",
              "prefillTokensPerSec": 2619.02,
              "decodeTokensPerSec": 190.61,
              "decodeTokens": 84,
              "caveats": [
                "84 decode tokens < 256; compare in-trial engine timings instead"
              ]
            }
          ],
          "confidence": "short-admission-probes"
        },
        "allRecorded": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 83,
          "wallDurationMs": {
            "n": 83,
            "median": 370664,
            "min": 259153,
            "max": 4998923
          },
          "successfulWallDurationMs": {
            "n": 74,
            "median": 359586.5,
            "min": 259153,
            "max": 4809732
          },
          "decodeTokensPerSec": {
            "n": 82,
            "median": 125.505,
            "min": 36.7,
            "max": 176.38
          },
          "prefillTokensPerSec": {
            "n": 82,
            "median": 2105.575,
            "min": 444.45,
            "max": 2774.2
          },
          "peakRssMb": {
            "n": 83,
            "median": 33149,
            "min": 916,
            "max": 40572
          },
          "toolCalls": {
            "n": 83,
            "median": 22,
            "min": 0,
            "max": 375
          }
        },
        "comparisonObservations": {
          "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
          "recorded": 82,
          "wallDurationMs": {
            "n": 82,
            "median": 370518.5,
            "min": 259153,
            "max": 4998923
          },
          "successfulWallDurationMs": {
            "n": 74,
            "median": 359586.5,
            "min": 259153,
            "max": 4809732
          },
          "decodeTokensPerSec": {
            "n": 81,
            "median": 126.21,
            "min": 36.7,
            "max": 176.38
          },
          "prefillTokensPerSec": {
            "n": 81,
            "median": 2117.17,
            "min": 608.21,
            "max": 2774.2
          },
          "peakRssMb": {
            "n": 82,
            "median": 32782,
            "min": 916,
            "max": 39543
          },
          "toolCalls": {
            "n": 82,
            "median": 21.5,
            "min": 0,
            "max": 375
          }
        },
        "bySuite": {
          "core": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 22,
            "wallDurationMs": {
              "n": 22,
              "median": 290987,
              "min": 259499,
              "max": 508466
            },
            "successfulWallDurationMs": {
              "n": 21,
              "median": 287926,
              "min": 259499,
              "max": 508466
            },
            "decodeTokensPerSec": {
              "n": 22,
              "median": 143.88,
              "min": 108.58,
              "max": 176.38
            },
            "prefillTokensPerSec": {
              "n": 22,
              "median": 2357.775,
              "min": 2054.95,
              "max": 2774.2
            },
            "peakRssMb": {
              "n": 22,
              "median": 29403,
              "min": 28354,
              "max": 36703
            },
            "toolCalls": {
              "n": 22,
              "median": 8.5,
              "min": 2,
              "max": 25
            }
          },
          "productivity": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 26,
            "wallDurationMs": {
              "n": 26,
              "median": 393083,
              "min": 259153,
              "max": 4998923
            },
            "successfulWallDurationMs": {
              "n": 23,
              "median": 370373,
              "min": 259153,
              "max": 1531616
            },
            "decodeTokensPerSec": {
              "n": 26,
              "median": 115.7,
              "min": 56.54,
              "max": 170.56
            },
            "prefillTokensPerSec": {
              "n": 26,
              "median": 1918.225,
              "min": 694.52,
              "max": 2714.82
            },
            "peakRssMb": {
              "n": 26,
              "median": 33082.5,
              "min": 28354,
              "max": 39543
            },
            "toolCalls": {
              "n": 26,
              "median": 26.5,
              "min": 4,
              "max": 375
            }
          },
          "developer": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 20,
            "wallDurationMs": {
              "n": 20,
              "median": 456504,
              "min": 265728,
              "max": 1458789
            },
            "successfulWallDurationMs": {
              "n": 18,
              "median": 456504,
              "min": 265728,
              "max": 1251587
            },
            "decodeTokensPerSec": {
              "n": 20,
              "median": 121.155,
              "min": 36.7,
              "max": 148.9
            },
            "prefillTokensPerSec": {
              "n": 20,
              "median": 2017.04,
              "min": 608.21,
              "max": 2619.43
            },
            "peakRssMb": {
              "n": 20,
              "median": 33246,
              "min": 28704,
              "max": 37409
            },
            "toolCalls": {
              "n": 20,
              "median": 25.5,
              "min": 3,
              "max": 92
            }
          },
          "complex-work": {
            "population": "unchanged fixed-eligible observations; reported comparison scope is explicit",
            "recorded": 14,
            "wallDurationMs": {
              "n": 14,
              "median": 511066,
              "min": 263688,
              "max": 4809732
            },
            "successfulWallDurationMs": {
              "n": 12,
              "median": 457598.5,
              "min": 263688,
              "max": 4809732
            },
            "decodeTokensPerSec": {
              "n": 13,
              "median": 100.33,
              "min": 45.79,
              "max": 140.82
            },
            "prefillTokensPerSec": {
              "n": 13,
              "median": 1675.36,
              "min": 814.58,
              "max": 2539.19
            },
            "peakRssMb": {
              "n": 14,
              "median": 36920.5,
              "min": 916,
              "max": 39260
            },
            "toolCalls": {
              "n": 14,
              "median": 49,
              "min": 0,
              "max": 123
            }
          }
        }
      }
    }
  },
  "modelMetadata": {
    "lfm2.5-24b-a2b-q4": {
      "label": "LFM2.5 (24B-A2B, Q4)",
      "tier": "medium",
      "parameterSize": "24B-A2B",
      "quantization": "Q4_0",
      "catalogVersion": "1.0.0",
      "weightHashes": {
        "LFM2.5-24B-A2B-QAD-Q4_0.gguf": "d1c03f73ecaf61017d70419eec9e5dc0b05d50ae26c23d522829b934fbabf8a5"
      }
    },
    "qwen3.6-35b-a3b-q4": {
      "label": "Qwen 3.6 (35B-A3B, Q4)",
      "tier": "medium",
      "parameterSize": "35B-A3B",
      "quantization": "UD-Q4_K_M",
      "catalogVersion": "1.0.3",
      "weightHashes": {
        "Qwen3.6-35B-A3B-UD-Q4_K_M.gguf": "0b21525e972670ed59e1812e170b27c26355381f0656ecc4e25617ece7dac58b"
      }
    },
    "qwen3.8-27b-q4": {
      "label": "Qwen 3.8 (27B, Q4)",
      "tier": "medium",
      "parameterSize": "27B",
      "quantization": "UD-Q4_K_M",
      "catalogVersion": "1.0.5",
      "weightHashes": {
        "mmproj-F16.gguf": "cbb841a9ee0636b2ec172f5bb8df2ea8dfeb01e90fe7c6126581d662a0b4e43e",
        "Qwen3.8-27B-UD-Q4_K_M.gguf": "322e194ff79741c7baa497c240f677f54b201b0efab44ca8e50f122b39123482"
      }
    },
    "btl4-35b-q4": {
      "label": "BTL-4 (35B-A3B, Q4)",
      "tier": "medium",
      "parameterSize": "35B-A3B",
      "quantization": "Q4_K_M",
      "catalogVersion": "1.0.1",
      "weightHashes": {
        "badtheorylabs_BTL-4-Q4_K_M.gguf": "75ce22e4f603bdf3f94c2b32151bb6d3492d2ab8a34ad8a4bfa2eaef0125c50f"
      }
    },
    "ornith1.5-35b-a3b-q4": {
      "label": "Ornith 1.5 (35B-A3B, Q4)",
      "tier": "medium",
      "parameterSize": "35B-A3B",
      "quantization": "Q4_K_M",
      "catalogVersion": "1.0.0",
      "weightHashes": {
        "Ornith-1.5-35B-Q4_K_M.gguf": "42739874cc2ccfdb8523b23fbe52e29b2a7555c8176737ca9ca0b5d59859d41f"
      }
    }
  },
  "enginePins": {
    "baseline": {
      "engine": "llama-cpp",
      "revision": "d81235049384534c167caea52b85a694f6103d14",
      "build": "11429",
      "backend": "metal"
    },
    "minicpmExperimental": {
      "revision": "69f201a2051ea9b9e9b50c3cc56afd9e2ae64414",
      "recordedTrials": 0,
      "note": "Experimental MiniCPM candidate was not evaluated; do not treat it as the baseline or publish a MiniCPM measurement."
    }
  },
  "caveats": {
    "issues": [
      {
        "id": "ab-test-readout-judge-rubric",
        "kind": "grader",
        "status": "documented-no-source-edits",
        "scenario": "craftbook-ab-test-readout",
        "reason": "Declared safety/mapping/signals/containment axes evaluate mock-service execution rather than the statistical readout. Affected quality judgment is withheld; deterministic pass/fail unchanged."
      },
      {
        "id": "tankcombat-runtime-coverage",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr",
        "reason": "Automated PASS does not validate combat/scoring: source never tests player-bullet collisions, keeps score at zero, and uses a 1200-second enemy firing interval. Inline quality judgment exposes the gap.",
        "additionalTrialIds": [
          "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb",
          "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn"
        ],
        "additionalEvidence": [
          {
            "trialId": "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb",
            "deterministicSuccess": true,
            "finding": "Ornith first repetition passes the frozen runtime checks but never marks the player as isPlayer. Every player shot is hostile; unmodified real-time replay loses three lives and reaches game over at score zero within 1.1 seconds of firing. The deterministic pass and fixed scores are preserved; inline functional completeness records the defect.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tankcombat-review-ornith-r1-2026-10-21T1011.json"
          },
          {
            "trialId": "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn",
            "deterministicSuccess": true,
            "finding": "Ornith repetition 2 passes frozen runtime checks, but collision detection is never invoked, score cannot increase and player bullet updates cancel to zero movement. Local real-time replay confirms stationary shots; enemies cannot be defeated. Fixed pass retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tankcombat-review-ornith-r2-2026-10-21T1957.json"
          }
        ]
      },
      {
        "id": "generated-gate-failure-attribution",
        "kind": "grader",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72",
        "reason": "Frozen failure classifier excludes an unevaluable {{workPath}} gate in a model-created recipe as infrastructure. Final analysis should distinguish framework interpolation faults from recipe-authoring defects so exclusions do not hide task errors. Original classification and trial score remain unchanged.",
        "additionalTrialIds": [
          "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4",
          "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc",
          "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1"
        ],
        "recoveryReviews": [
          "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/recovery-reviews/2026-10-18T11-03-37-484Z/review.json",
          "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/recovery-reviews/2026-10-21T1612/review.json",
          "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/recovery-reviews/2026-10-21T1842/review.json"
        ],
        "additionalEvidence": [
          {
            "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge",
            "finding": "Generated task references missing inventoryGate in craftbook scope; later script still lacks gateResult(...) invocation. Frozen infrastructure exclusion preserved; audit permits continuation in fresh workspaces without changing source.",
            "recoveryReview": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/recovery-reviews/2026-10-20T13-07-10-478Z/review.json"
          },
          {
            "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc",
            "finding": "Ornith generated a supplier-intake gate with {{params.reportPath}}/{{params.cleanPath}} despite launch keys reportPath/cleanPath. Gate pauses as unevaluable. Complete 33085-line, zero-gap capture reconciles 84 telemetry tool calls, including two gate failures omitted from a persisted session. Frozen infrastructure exclusion retained; fresh identities match and continuation uses new trial homes.",
            "recoveryReview": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/recovery-reviews/2026-10-21T1612/review.json"
          },
          {
            "trialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1",
            "finding": "Ornith generated checkInventoryReportGate imports defineScript/gezel from the checks-only SDK entry. The gate pauses unevaluably; model loops until the existing 18-minute plateau guard terminates the trial. Frozen infrastructure exclusion preserved. 41231 zero-gap events reconcile all 78 telemetry tools; twelve persisted-session omissions and controlled mid-turn cancellation are audited. Fresh shared/full identities match.",
            "recoveryReview": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/recovery-reviews/2026-10-21T1842/review.json"
          }
        ]
      },
      {
        "id": "week-plan-calendar-provenance",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn",
        "deterministicSuccess": false,
        "reason": "Completed week-plan artifact contains a fully populated calendar and claimed meeting moves while mock-requests.json has calendar: []. Audit fixture-call provenance and actual event grounding in the final report; deterministic outcome remains unchanged.",
        "additionalTrialIds": [
          "craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj",
          "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v",
          "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0",
          "craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8",
          "craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4",
          "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z",
          "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u"
        ],
        "additionalEvidence": [
          {
            "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj",
            "finding": "Model failure retained: report is in artifacts instead of required workspace path. Calendar calls are empty; report admits failed retrieval but invents nine October meetings against the eight-event July fixture, and misses the actual Q3 budget/Priya overlap."
          },
          {
            "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v",
            "finding": "Second repetition preserves timeout/model-default. Required report is in artifacts instead of workspace. Calendar calls are empty; report admits no retrieval, invents eleven October meetings against eight-event July fixture, misses budget/Priya overlap, and schedules SOW after its invented demo even in its alternative fix."
          },
          {
            "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0",
            "deterministicSuccess": true,
            "finding": "Successful probe at 14:32:03Z returned week 2026-07-13 and eight events, but the final report was written at 14:31:00Z and still uses Oct 19–23, daily standups and invented Sam/Meridian/Product/Onboarding meetings. It omits the actual Tue 10:00 budget/Priya 10:30 conflict; Thu 10–13 focus overlaps Design review 9:30–10:30, Fri 10–12 focus overlaps retro 11–12. Real retrieval did not cause a content-grounding correction.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/week-plan-review-qwen38-r3-2026-10-19T1445.json"
          },
          {
            "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8",
            "deterministicSuccess": false,
            "finding": "Two successful authenticated calendar probes returned all eight July events at 03:07:37Z and 03:08:28Z. The final report claims the service was unreachable and calendar empty, misses all events, both prep meetings and the Tuesday conflict, and places 09–11 focus blocks across real meetings.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/week-plan-review-btl4-r1-2026-10-20T0332.json"
          },
          "evals/runs/scorecard-seven-models-2026-10-15-proposal/week-plan-review-btl4-r3-2026-10-21T0311.json",
          {
            "trialId": "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z",
            "deterministicSuccess": false,
            "finding": "No calendar probe ran; the required workspace output is absent. The produced drawer report invents most meetings, misses the actual budget/Priya overlap and schedules focus across real meetings. Existing task failure retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/week-plan-review-ornith-r1-2026-10-21T1226.json"
          },
          {
            "trialId": "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u",
            "deterministicSuccess": false,
            "finding": "The required workspace report is absent; a drawer report honestly leaves all calendar-derived sections pending. No authenticated calendar request ran. All eight meetings, agenda-less prep needs, the Tuesday overlap and focus blocks are missing. Existing model-default failure and fixed scores are retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/week-plan-review-ornith-r2-2026-10-21T2327.json"
          }
        ]
      },
      {
        "id": "deep-security-safe-section-label",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1",
        "deterministicSuccess": false,
        "reason": "Failure evidence requires the literal section text Verified Safe or Not Statically Verifiable, while the produced report has a Reviewed and confirmed correct (no findings) section explicitly covering the seeded safe patterns. Audit label matching versus substantive review quality in the final analysis; no claim that this is the only failing check. Model-default classification and all scores remain unchanged.",
        "additionalTrialIds": [
          "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1",
          "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k"
        ],
        "additionalEvidence": [
          {
            "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1",
            "deterministicSuccess": false,
            "finding": "Third repetition retains model-default/stale no-write failure at 9/11. Required literal Verified Safe or Not Statically Verifiable is absent, while Reviewed and Correct (no findings) substantively clears profile authorization/placeholders, static shell and allow-listed ORDER BY. This does not assert the label is the only unmet check; original fixed score and failure stay unchanged."
          },
          {
            "trialId": "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k",
            "deterministicSuccess": false,
            "finding": "Safe constructs (not findings) explicitly clears the seeded safe patterns, but the frozen gate requires literal Verified Safe or Not Statically Verifiable. This is only one of three unmet checks: systemic analysis and JSON theme fields are genuinely missing. Original 8/11 model-default failure and fixed scores remain unchanged.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/security-review-ornith-r2-2026-10-22T0042.json"
          }
        ]
      },
      {
        "id": "invoice-run-books-not-updated",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i",
        "deterministicSuccess": true,
        "reason": "Frozen gate passes three correctly numbered invoice HTML files and a run summary, but captured ledger.json retains lastInvoiceNumber 2026-041 and the three billed entries remain invoiced:false. Inline completeness score records the missing books update; final analysis should distinguish invoice production from durable bookkeeping. Deterministic outcome and fixed scores are unchanged.",
        "additionalTrialIds": [
          "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs",
          "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s",
          "craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk",
          "craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x",
          "craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd",
          "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p",
          "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v"
        ],
        "additionalEvidence": [
          {
            "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs",
            "finding": "Second repetition also passes the frozen gate while ledger remains at 041 with current entries uninvoiced. Invoice due dates use work dates; original issue date was removed rather than fixing net terms.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-qwen38-r2-2026-10-19T0644.json"
          },
          {
            "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s",
            "deterministicSuccess": true,
            "finding": "Third repetition also passes while ledger.json retains lastInvoiceNumber 2026-041 and all three billed entries remain invoiced:false. This time issue-date-based net terms are correct, but report EUR conflicts with first invoice USD and others omit currency; July services are described as October. Original fixed outcome unchanged.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-qwen38-r3-2026-10-19T1917.json"
          },
          {
            "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk",
            "deterministicSuccess": true,
            "finding": "Books unchanged at 041 and all three false flags. Invoice 043 additionally fabricates bank details, names the client as payee, adds late interest and euro currency against a dollar report; other invoices invent addresses/contacts.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-btl4-r1-2026-10-20T1309.json"
          },
          {
            "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x",
            "deterministicSuccess": true,
            "finding": "BTL4 second repetition also passes while ledger retains 041 and all three current entries remain uninvoiced. Amounts/numbers correct, but invoices 042/044 invent November issue dates with 14-day footer terms for net-30 clients; unsupported sender/address/contact details. All three print on one A4 page with excessive top whitespace. Fixed pass and scores retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-btl4-r2-2026-10-20T2056.json"
          },
          {
            "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd",
            "deterministicSuccess": true,
            "finding": "BTL4 third repetition also leaves books at 041 with all three billed entries uninvoiced. Correct amounts/sequence but unsupported sent claim, missing visible invoice number/date on 042, incorrect 043/044 terms and fabricated contacts. Saved reviewer marks terms FAIL; final invoices remain unrepaired. Print review shows two, two and one pages. Frozen pass and scores preserved.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-btl4-r3-2026-10-21T0626.json"
          },
          {
            "trialId": "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p",
            "deterministicSuccess": true,
            "finding": "Ornith r1 also leaves books at 041 with all three billed entries uninvoiced. Correct amounts/sequence but unsupported sending claim, future Nov 1 issue dates conflicting with due dates/terms, invented contacts and wrong client header on 044. All three print cleanly on one Letter page. Frozen pass and scores retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-ornith-r1-2026-10-21T1512.json"
          },
          {
            "trialId": "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v",
            "deterministicSuccess": true,
            "finding": "Ornith r2 passes while ledger remains at 041 and all three current entries remain uninvoiced. Correct totals, sequence and issue-date-based due dates; summary dispatch claim is unsupported. All three fit one A4 page, but 042/043 print gray lower blocks and 044 totals are misaligned. Fixed pass and scores retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-ornith-r2-2026-10-22T0157.json"
          }
        ]
      },
      {
        "id": "spreadsheet-model-month-column-indexing",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh",
        "deterministicSuccess": true,
        "reason": "Frozen processability gate passes the full workflow, but local browser review confirms global i%12 indexing across 13-cell rows shifts the monthly revenue/COGS/profit columns. M1 displays $4,609 revenue/$1,788 COGS/$2,606 gross profit instead of $4,760/$1,904/$2,856. Headline totals and flat sanity scenario work. Captured reviewer hand arithmetic also asserts incorrect defaults ($48,434 revenue/$29,061 GP versus actual $49,393/$29,636). Original deterministic outcome and fixed scores remain unchanged.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-qwen38-r2-2026-10-19T0030.json"
      },
      {
        "id": "petshop-html-first-order",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q",
        "deterministicSuccess": true,
        "reason": "Frozen prompt explicitly requires shipping HTML with assets/logo.png wired in before image generation. Captured tool events show generate_image at 12:12:52Z and first index.html write at 12:15:35Z, yet deterministic trial passes. Final analysis should separate resulting file correctness from required execution order; fixed outcome remains unchanged.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/petshop-review-qwen38-r3-2026-10-19T1217.json",
        "additionalTrialIds": [
          "petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4",
          "petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex"
        ],
        "additionalEvidence": [
          {
            "trialId": "petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4",
            "deterministicSuccess": true,
            "finding": "Ornith first repetition generated the logo at 10:03:46 before writing shipping HTML at 10:04:35. The earlier report.md is not index.html. Final same-workspace logo works, but the explicit HTML-first order was not followed. Fixed pass retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/petshop-review-ornith-r1-2026-10-21T1011.json"
          },
          {
            "trialId": "petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex",
            "deterministicSuccess": true,
            "finding": "Ornith repetition 2 generates its logo at 19:51:45 before the sole shipping HTML write at 19:52:26. Same-workspace PNG works, but explicit HTML-first sequencing is not met. Fixed pass retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/petshop-review-ornith-r2-2026-10-21T1957.json"
          }
        ]
      },
      {
        "id": "petshop-missing-companion-assets",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv",
        "deterministicSuccess": true,
        "reason": "Frozen petshop gate passed HTML plus generated PNG, but shipping index.html references styles.css and app.js that were never captured or written. Local rendering confirms unstyled page, empty dynamic content and inert cart. Inline quality records incomplete delivery; fixed scores and success remain unchanged.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/petshop-review-btl4-r1-2026-10-20T0118.json"
      },
      {
        "id": "spreadsheet-model-false-verification",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k",
        "deterministicSuccess": true,
        "reason": "Frozen runtime gates passed despite scope/evaluate claiming verified net profit 1375 while browser computes 20980, a dimensionally incorrect effective-CAC formula, and a missing-total runtime exception truncating output rows. Inline review records these defects; fixed capability outcome and scores unchanged.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-btl4-r1-2026-10-20T0233.json",
        "additionalTrialIds": [
          "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k"
        ],
        "additionalEvidence": [
          {
            "trialId": "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k",
            "deterministicSuccess": true,
            "finding": "Ornith repetition 2 build note falsely verifies scope net82752/payback8.77. Independent local replay gives33001.36/21.82. Reviewer catches the contract error and says it remains blocked, but stale-step edits fail and the task reaches Finish without correcting the saved scope. Frozen PASS retained.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-ornith-r2-2026-10-21T2142.json"
          }
        ]
      },
      {
        "id": "bug-fix-tdd-original-tests-overwritten",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme",
        "deterministicSuccess": true,
        "reason": "Boundary reproduction overwrote all three original seeded tests. Fixed gate and mutant guard pass, but original small-order, large-order and invalid-quantity coverage vanished. Inline scores reflect reduced verification and incomplete touched-file disclosure; preserve fixed pass/fail.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tdd-review-btl4-r1-2026-10-20T0602.json",
        "additionalTrialIds": [
          "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m"
        ],
        "additionalEvidence": [
          {
            "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m",
            "deterministicSuccess": false,
            "finding": "Second repetition again overwrites the original three tests, later replacing them with exact/below/above threshold checks. Real red/green receipts substantiate the fix, but invalid-quantity and original broad coverage disappear; notes claim unchanged tests and omit test-file edit. Trial ultimately fails drawer fix-notes sections because repairs target workspace copy. Fixed outcome and scores unchanged.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tdd-review-btl4-r2-2026-10-20T1948.json"
          }
        ]
      },
      {
        "id": "spreadsheet-recurring-opex-payback",
        "kind": "grader-coverage-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w",
        "deterministicSuccess": true,
        "reason": "Scope and live calculator label recurring monthly opex/gross savings as months to payback despite permanently negative net savings; acceptance checklist contradicts its own net calculation and strict float equality disables default sanity check. Inline quality reflects these errors; fixed pass/fail remains unchanged.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-btl4-r2-2026-10-20T1541.json",
        "additionalTrialIds": [
          "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89",
          "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k"
        ],
        "additionalEvidence": [
          "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-btl4-r3-2026-10-21T0056.json",
          {
            "trialId": "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k",
            "deterministicSuccess": true,
            "finding": "Ornith repetition 2 calls an annual staffing expense the payback capital after already subtracting it in annual net savings. No one-time investment is defined. Source also treats first-response delay as saved labor time and falsely maps CSV revenue mean1190 to per-SKU monthly70.",
            "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-ornith-r2-2026-10-21T2142.json"
          }
        ]
      },
      {
        "id": "memory-admission-model-default-attribution",
        "kind": "failure-attribution-gap",
        "status": "documented-no-source-edits",
        "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
        "deterministicSuccess": false,
        "fixedFailureClass": "model",
        "fixedFailureClassRule": "model-default",
        "reason": "The fixed model-default eligibility remains untouched. With zero engine launches, no first token and zero tools, the captured failure does not establish model task capability. Memory admission is an advisory environmental-attribution caveat for the final comparison, not a newly fabricated exclusion.",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/memory-admission-observation-2026-10-22T0257.json"
      }
    ],
    "memoryAdmission": {
      "checkedAt": "2026-10-22T03:01:09.835638Z",
      "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
      "runDir": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
      "status": "completed-unaltered-model-failure-driver-continued",
      "intervention": null,
      "observations": [
        "At 02:54:19Z and 02:54:24Z, initial and harness recovery turns both failed admission before any native engine launched. Daemon recorded 23.2–23.6 GB live free RAM and only 18.7–19.1 GB safely available against 22.5 GB estimated weights.",
        "Host memory was substantially freer by 02:58:32Z: 3,671,534 free 16,384-byte pages (~56.0 GiB), memory_pressure -Q reported 64% free. No foreign native engine was present in the host process snapshot. Cause of the transient pressure is not established.",
        "Driver, runner and daemon remain alive; no stopped marker or completed result exists yet. No retry, nudge, process termination, tuning or watchdog change was made by this monitor."
      ],
      "captureFiles": {
        "daemon.log": {
          "bytes": 9539,
          "sha256": "6d3450c3a667d29d8b6631112ad4f49046d166d50aa41b958ed455ed3f56ed7d",
          "modifiedAt": "2026-10-22T02:59:22.280532+00:00"
        },
        "log.txt": {
          "bytes": 6608,
          "sha256": "4f24ad9e0f9cd880e2348fd28605e7feb3dc7ca9471671ab6cdd390f72370710",
          "modifiedAt": "2026-10-22T02:59:22.340618+00:00"
        },
        "recording/chat-events.jsonl": {
          "bytes": 7679,
          "sha256": "8052a8a494837016cc8726e425f56ee61c7ec338254daa45852f8f1fd5a527b9",
          "modifiedAt": "2026-10-22T02:56:19.175147+00:00"
        }
      },
      "daemonDenials": [
        "2026-10-22T02:54:19.303Z INFO  [chat] [llama-cpp] ornith1.5-35b-a3b-q4: kv=q8_0 (f16 does not fit the memory plan — q8_0 for the memory saving)",
        "2026-10-22T02:54:19.467Z WARN  [chat] [llama-cpp] Not enough memory to run Ornith 1.5 (35B-A3B, Q4) with its required 65,536-token working window. Even at one engine slot, this machine could safely admit only about 1 tokens per turn. context clamped 65536 → 1 tokens/turn (1 slot): weights ~22.5 GB + KV at the requested context ~0.7 GB exceeds available memory ~19.0 GB (budget 112.0 GB, live free RAM 23.5 GB); KV now ~0.0 GB. Sessions compact sooner instead of the machine paging.",
        "2026-10-22T02:54:19.468Z WARN  [engine-router] bindForSession: prior key llama-cpp:ornith1.5-35b-a3b-q4:0 unavailable (There is not enough memory to run this model. Please select another model or close some applications on this device.); rebinding to least-loaded replica",
        "2026-10-22T02:54:19.472Z INFO  [chat] [llama-cpp] ornith1.5-35b-a3b-q4: kv=q8_0 (f16 does not fit the memory plan — q8_0 for the memory saving)",
        "2026-10-22T02:54:19.475Z WARN  [chat] [llama-cpp] Not enough memory to run Ornith 1.5 (35B-A3B, Q4) with its required 65,536-token working window. Even at one engine slot, this machine could safely admit only about 1 tokens per turn. context clamped 65536 → 1 tokens/turn (1 slot): weights ~22.5 GB + KV at the requested context ~0.7 GB exceeds available memory ~19.1 GB (budget 112.0 GB, live free RAM 23.6 GB); KV now ~0.0 GB. Sessions compact sooner instead of the machine paging.",
        "There is not enough memory to run this model. Please select another model or close some applications on this device.",
        "2026-10-22T02:54:19.477Z DEBUG [chat] runSend#61cd71ad EXIT throw afterMs=189 There is not enough memory to run this model. Please select another model or close some applications on this device.",
        "2026-10-22T02:54:24.443Z INFO  [chat] [llama-cpp] ornith1.5-35b-a3b-q4: kv=q8_0 (f16 does not fit the memory plan — q8_0 for the memory saving)",
        "2026-10-22T02:54:24.449Z WARN  [chat] [llama-cpp] Not enough memory to run Ornith 1.5 (35B-A3B, Q4) with its required 65,536-token working window. Even at one engine slot, this machine could safely admit only about 1 tokens per turn. context clamped 65536 → 1 tokens/turn (1 slot): weights ~22.5 GB + KV at the requested context ~0.7 GB exceeds available memory ~18.7 GB (budget 112.0 GB, live free RAM 23.2 GB); KV now ~0.0 GB. Sessions compact sooner instead of the machine paging.",
        "2026-10-22T02:54:24.449Z WARN  [engine-router] bindForSession: prior key llama-cpp:ornith1.5-35b-a3b-q4:0 unavailable (There is not enough memory to run this model. Please select another model or close some applications on this device.); rebinding to least-loaded replica",
        "2026-10-22T02:54:24.452Z INFO  [chat] [llama-cpp] ornith1.5-35b-a3b-q4: kv=q8_0 (f16 does not fit the memory plan — q8_0 for the memory saving)",
        "2026-10-22T02:54:24.456Z WARN  [chat] [llama-cpp] Not enough memory to run Ornith 1.5 (35B-A3B, Q4) with its required 65,536-token working window. Even at one engine slot, this machine could safely admit only about 1 tokens per turn. context clamped 65536 → 1 tokens/turn (1 slot): weights ~22.5 GB + KV at the requested context ~0.7 GB exceeds available memory ~18.7 GB (budget 112.0 GB, live free RAM 23.2 GB); KV now ~0.0 GB. Sessions compact sooner instead of the machine paging.",
        "There is not enough memory to run this model. Please select another model or close some applications on this device.",
        "2026-10-22T02:54:24.457Z DEBUG [chat] runSend#61cd71ad EXIT throw afterMs=23 There is not enough memory to run this model. Please select another model or close some applications on this device.",
        "2026-10-22T02:54:24.457Z WARN  [http] [sessions] send failed for 61cd71ad-8d3a-49e0-bd6e-5ab1aadc5eb3: There is not enough memory to run this model. Please select another model or close some applications on this device."
      ],
      "latestTrialLog": [
        "[2026-10-22T02:57:26.061Z] [poll] poisoned-session recovery unanswered: reza/61cd71ad (182s ago) — repair turn dispatched, no turn in flight and no progress since; the session is not working on it",
        "[2026-10-22T02:58:26.559Z] [poll] poisoned-session recovery unanswered: reza/61cd71ad (242s ago) — repair turn dispatched, no turn in flight and no progress since; the session is not working on it",
        "[2026-10-22T02:59:21.979Z] [poll] no-progress (soft): no engine activity for 303s (threshold=300s); chat stalled — failing trial",
        "[2026-10-22T02:59:22.238Z] [perf] metrics: peakRss=916MB, gpu=n/a",
        "[2026-10-22T02:59:22.243Z] [recording] tap stopped: 18 line(s), 0 coalesced delta(s), 0 gap(s)",
        "[2026-10-22T02:59:22.247Z] [capture] dumped 1 session(s)",
        "[2026-10-22T02:59:22.248Z] [capture] snapshotted telemetry for 0 session(s)",
        "[2026-10-22T02:59:22.259Z] [capture] snapshotted artifacts/workspace/history for 3 project(s)",
        "[2026-10-22T02:59:22.265Z] [recording] captured notes for 2 task(s)",
        "[2026-10-22T02:59:22.272Z] [recording] captured 5 actor(s)",
        "[2026-10-22T02:59:22.274Z] [distill] wrote recording/transcript.json (4 scenes, 5510 bytes)",
        "[2026-10-22T02:59:22.275Z] [recording] wrote recording/manifest.json",
        "[2026-10-22T02:59:22.338Z] [trial] daemon shut down",
        "[2026-10-22T02:59:22.340Z] [trial] FAIL duration=542147ms reason=chat stalled for 303s — daemon reachable but issuing no model turns (soft digest stuck at 7a15590c247174ba)"
      ],
      "result": {
        "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
        "scenarioId": "craftbook-export-generalize",
        "modelId": "ornith1.5-35b-a3b-q4",
        "modelTier": "medium",
        "generalistMode": "auto",
        "engine": "llama-cpp",
        "startedAt": "2026-10-22T02:50:20.192Z",
        "finishedAt": "2026-10-22T02:59:22.339Z",
        "durationMs": 542147,
        "success": false,
        "reason": "chat stalled for 303s — daemon reachable but issuing no model turns (soft digest stuck at 7a15590c247174ba)",
        "runDir": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
        "failureMode": "chat-stalled",
        "engineContext": {
          "launches": 0,
          "clamp": {
            "requestedTokens": 65536,
            "grantedTokens": 1
          }
        },
        "finalSniff": {
          "key": "craftbook-export-generalize",
          "score": 1,
          "bytes": 0,
          "deliverableMissing": true,
          "failReason": "out/q1-intake.md does not exist yet — write the short Q1 intake report there: what you checked and which suppliers have an impossible lead time"
        },
        "failureClass": "model",
        "failureClassRule": "model-default"
      },
      "eligibility": {
        "failureClass": "model",
        "includedInModelAggregate": true,
        "failureClassRule": "model-default",
        "evidence": [
          {
            "source": "result",
            "path": "failureClass",
            "value": "model"
          }
        ]
      },
      "captureManifest": {
        "schemaVersion": 1,
        "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
        "scenarioId": "craftbook-export-generalize",
        "modelId": "ornith1.5-35b-a3b-q4",
        "startedAt": "2026-10-22T02:50:20.192Z",
        "finishedAt": "2026-10-22T02:59:22.274Z",
        "files": {
          "chatEvents": "recording/chat-events.jsonl",
          "taskNotes": "recording/task-notes.json",
          "actors": "recording/actors.json",
          "transcript": "recording/transcript.json"
        },
        "chatEvents": {
          "lines": 18,
          "coalescedDeltas": 0
        },
        "capture": {
          "taskNotes": "ok",
          "taskHistory": "absent",
          "actors": "ok",
          "screenshots": "absent"
        }
      },
      "fixedFileSha256": {
        "result.json": "c98d7358fd46b43ed5fa5863270402dfce020251bb20e0e1c2a19d64565d8636",
        "score.json": "1e0d9e8712c7cc2a31034e6856956b4ae5d8ef50c418b6a9c88f92ce7f343890",
        "facts.json": "71acb722a4a55666ddae457d423d5d2c037af3f6687c65feee6421a7f71610e1",
        "metrics.json": "66e629fb789df0a406526eee4c4e60e96f30cac1806b5574dc1b8bc469f32045"
      },
      "nextCell": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params",
      "stopped": null,
      "nextTrialId": "craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy",
      "processObservation": "At 03:00:13Z, failed trial runner 43767 and daemon 44795 were absent; driver 56920 / lease 56919 remained alive and new params runner 48589 / lease 48588 with copy child 48665 was alive. No native engine or duplicate runner observed.",
      "interpretation": "The fixed model-default eligibility remains untouched. With zero engine launches, no first token and zero tools, the captured failure does not establish model task capability. Memory admission is an advisory environmental-attribution caveat for the final comparison, not a newly fabricated exclusion.",
      "followup": {
        "checkedAt": "2026-10-22T03:13:39.230055Z",
        "status": "fresh-next-trial-admitted-and-making-real-progress",
        "trialId": "craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy",
        "processes": {
          "driver": 56920,
          "lease": 56919,
          "runner": 48589,
          "runnerLease": 48588,
          "daemon": 50165,
          "engine": 50512
        },
        "admissionEvidence": [
          "2026-10-22T03:03:47.431Z INFO  [chat] [llama-cpp] ornith1.5-35b-a3b-q4: kv=f16 (f16 fits the full plan)",
          "2026-10-22T03:03:47.635Z INFO  [chat] [llama-cpp] ornith1.5-35b-a3b-q4: growing context 65536 → 262144 tokens/turn at 1 slot (target 262144): weights ~22.5 GB + grown KV ~5.0 GB fits fast memory ~112.0 GB",
          "2026-10-22T03:03:47.639Z INFO  [chat] [llama-cpp] offload plan (ornith1.5-35b-a3b-q4): MoE fits VRAM (weights 20.2 GiB + reserves 6.5 GiB ≤ 107.5 GiB) — full GPU residency",
          "2026-10-22T03:04:36.783Z INFO  [llama-cpp] [llama-cpp] TTFT 47758ms (session model=ornith1.5-35b-a3b-q4)"
        ],
        "latestNativeProgress": [
          "2026-10-22T03:13:32.782Z INFO  [chat] [llama-server] 9.43.670.373 I slot print_timing: id  0 | task 5147 | prompt processing, n_tokens =  54449, progress = 1.00, t =  67.35 s / 808.50 tokens per second",
          "2026-10-22T03:13:35.927Z INFO  [chat] [llama-server] 9.46.815.212 I slot print_timing: id  0 | task 5147 | n_gen =    218, tg =  71.40 t/s, tg_3s =  71.73 t/s",
          "2026-10-22T03:13:38.973Z INFO  [chat] [llama-server] 9.49.861.308 I slot print_timing: id  0 | task 5147 | n_gen =    396, tg =  64.91 t/s, tg_3s =  58.44 t/s"
        ],
        "latestHarnessProgress": [
          "[2026-10-22T03:12:24.954Z] [poll] hard-progress: de4c8d73aa4e4961 → 806988850ca092b7 after 137s (workspaceProjects=3 sessions=3 daemon[src=service-telemetry](turns=3,tools=26,slot=1024,stream=17907) sniff=5/50B)"
        ],
        "interpretation": "Fresh next-cell inference confirms the earlier memory-admission failure was not persisting at this check. No monitor restart, retry or input change was needed. The prior recorded cell and advisory classification caveat remain preserved."
      }
    },
    "comparisonDrift": {
      "affectedTrialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st",
      "expectedSourceHash": "d68acf8def595c54deea3468ab9efe7ff2c6970be4e1d64e6fe034dbb82dd042",
      "observedSourceHash": "dc77f554edb4b1b718b6c2dccfc05003e18bea6555b1c6d54b22b6c4bb1cae3b",
      "sharedChanges": [
        {
          "component": "fixtures",
          "paths": [
            "auto-answer.ts",
            "batch.ts",
            "bin/ab-work-routing.ts",
            "bin/image-bench.ts",
            "craftbooks/adapter.ts",
            "craftbooks/authoring/author-linear.ts",
            "craftbooks/authoring/edit-midtask.ts",
            "craftbooks/authoring/export-generalize.ts",
            "craftbooks/authoring/gate-script.ts",
            "craftbooks/authoring/helpers.ts",
            "craftbooks/authoring/route-multi.ts",
            "craftbooks/bundled-toolsets.ts",
            "craftbooks/scenario.ts",
            "craftbooks/specs.ts",
            "craftbooks/test-spec-loader.ts",
            "craftbooks/types.ts",
            "docblocks-runtime.ts",
            "engine-context.ts",
            "fixtures/office-documents.ts",
            "image-bench/IMAGE-BENCH.md",
            "image-bench/alignment.ts",
            "image-bench/judge.ts",
            "image-bench/memory.ts",
            "image-bench/prompts.ts",
            "image-bench/sanity.ts",
            "image-bench/sd-log.ts",
            "index-bench/LEVERS.md",
            "native-bin.ts",
            "runner.ts",
            "scenarios/conflict-synthesis.ts",
            "scenarios/docblocks-integration.ts",
            "scenarios/docblocks-theme-roundtrip.ts",
            "scenarios/first-projects.ts",
            "scenarios/helpers.ts",
            "scenarios/incident-postmortem.ts",
            "scenarios/index.ts",
            "scenarios/ops-runbook.ts",
            "scenarios/powerpoint-sources.ts",
            "scenarios/tool-provenance.ts",
            "scenarios/tool-routing-format.ts",
            "scenarios/work-routing.ts",
            "stale-step-session.ts",
            "success-check.ts",
            "suites.ts",
            "work-routing/analyze.ts",
            "work-routing/classify.ts",
            "work-routing/probes.ts",
            "work-routing/rescore.ts"
          ]
        },
        {
          "component": "runtime",
          "paths": [
            "@bendyline/gezel-mcp/package.json",
            "@bendyline/gezel-service/dependency/@bendyline/docblocks-cli",
            "@bendyline/gezel/dist/browser.js",
            "@bendyline/gezel/dist/chunk-3Z6PBVZM.js",
            "@bendyline/gezel/dist/chunk-5BXY37IX.js",
            "@bendyline/gezel/dist/chunk-6ASB7JVJ.js",
            "@bendyline/gezel/dist/chunk-6RY2SMVG.js",
            "@bendyline/gezel/dist/chunk-6YKM4W6M.js",
            "@bendyline/gezel/dist/chunk-7BFSCPZK.js",
            "@bendyline/gezel/dist/chunk-7QFCY42H.js",
            "@bendyline/gezel/dist/chunk-AOKPYNLV.js",
            "@bendyline/gezel/dist/chunk-B7GVWVXH.js",
            "@bendyline/gezel/dist/chunk-BB6SJOHT.js",
            "@bendyline/gezel/dist/chunk-CN3XHZA2.js",
            "@bendyline/gezel/dist/chunk-CZMCPJOM.js",
            "@bendyline/gezel/dist/chunk-DG6IPRC4.js",
            "@bendyline/gezel/dist/chunk-DUS3LUJM.js",
            "@bendyline/gezel/dist/chunk-HKEUN4VQ.js",
            "@bendyline/gezel/dist/chunk-KK2H55XN.js",
            "@bendyline/gezel/dist/chunk-KX3ESUXI.js",
            "@bendyline/gezel/dist/chunk-OY2IFYYQ.js",
            "@bendyline/gezel/dist/chunk-PF7H4CIR.js",
            "@bendyline/gezel/dist/chunk-PL6T3IO4.js",
            "@bendyline/gezel/dist/chunk-XJYBP6IZ.js",
            "@bendyline/gezel/dist/chunk-XTAG6E6T.js",
            "@bendyline/gezel/dist/chunk-YFNRCWMC.js",
            "@bendyline/gezel/dist/index.js",
            "@bendyline/gezel/dist/local-loop/index.js",
            "@bendyline/gezel/dist/markdown/index.js",
            "@bendyline/gezel/dist/native/index.js",
            "@bendyline/gezel/dist/paths.js",
            "@bendyline/gezel/dist/runtime/index.js",
            "@bendyline/gezel/dist/schemas/app-models.js",
            "@bendyline/gezel/dist/schemas/index.js",
            "@bendyline/gezel/dist/schemas/queue-status.js",
            "@bendyline/gezel/dist/speech-models/index.js"
          ]
        },
        {
          "component": "source",
          "paths": [
            "evals/src/auto-answer.ts",
            "evals/src/batch.ts",
            "evals/src/bin/ab-work-routing.ts",
            "evals/src/bin/image-bench.ts",
            "evals/src/craftbooks/adapter.ts",
            "evals/src/craftbooks/authoring/author-linear.ts",
            "evals/src/craftbooks/authoring/edit-midtask.ts",
            "evals/src/craftbooks/authoring/export-generalize.ts",
            "evals/src/craftbooks/authoring/gate-script.ts",
            "evals/src/craftbooks/authoring/helpers.ts",
            "evals/src/craftbooks/authoring/route-multi.ts",
            "evals/src/craftbooks/bundled-toolsets.ts",
            "evals/src/craftbooks/scenario.ts",
            "evals/src/craftbooks/specs.ts",
            "evals/src/craftbooks/test-spec-loader.ts",
            "evals/src/craftbooks/types.ts",
            "evals/src/docblocks-runtime.ts",
            "evals/src/engine-context.ts",
            "evals/src/fixtures/office-documents.ts",
            "evals/src/image-bench/IMAGE-BENCH.md",
            "evals/src/image-bench/alignment.ts",
            "evals/src/image-bench/judge.ts",
            "evals/src/image-bench/memory.ts",
            "evals/src/image-bench/prompts.ts",
            "evals/src/image-bench/sanity.ts",
            "evals/src/image-bench/sd-log.ts",
            "evals/src/index-bench/LEVERS.md",
            "evals/src/native-bin.ts",
            "evals/src/runner.ts",
            "evals/src/scenarios/conflict-synthesis.ts",
            "evals/src/scenarios/docblocks-integration.ts",
            "evals/src/scenarios/docblocks-theme-roundtrip.ts",
            "evals/src/scenarios/first-projects.ts",
            "evals/src/scenarios/helpers.ts",
            "evals/src/scenarios/incident-postmortem.ts",
            "evals/src/scenarios/index.ts",
            "evals/src/scenarios/ops-runbook.ts",
            "evals/src/scenarios/powerpoint-sources.ts",
            "evals/src/scenarios/tool-provenance.ts",
            "evals/src/scenarios/tool-routing-format.ts",
            "evals/src/scenarios/work-routing.ts",
            "evals/src/stale-step-session.ts",
            "evals/src/success-check.ts",
            "evals/src/suites.ts",
            "evals/src/work-routing/analyze.ts",
            "evals/src/work-routing/classify.ts",
            "evals/src/work-routing/probes.ts",
            "evals/src/work-routing/rescore.ts"
          ]
        }
      ],
      "comparisonChanges": [
        {
          "component": "fixtures",
          "paths": [
            "auto-answer.ts",
            "batch.ts",
            "bin/ab-work-routing.ts",
            "bin/image-bench.ts",
            "craftbooks/adapter.ts",
            "craftbooks/authoring/author-linear.ts",
            "craftbooks/authoring/edit-midtask.ts",
            "craftbooks/authoring/export-generalize.ts",
            "craftbooks/authoring/gate-script.ts",
            "craftbooks/authoring/helpers.ts",
            "craftbooks/authoring/route-multi.ts",
            "craftbooks/bundled-toolsets.ts",
            "craftbooks/scenario.ts",
            "craftbooks/specs.ts",
            "craftbooks/test-spec-loader.ts",
            "craftbooks/types.ts",
            "docblocks-runtime.ts",
            "engine-context.ts",
            "fixtures/office-documents.ts",
            "image-bench/IMAGE-BENCH.md",
            "image-bench/alignment.ts",
            "image-bench/judge.ts",
            "image-bench/memory.ts",
            "image-bench/prompts.ts",
            "image-bench/sanity.ts",
            "image-bench/sd-log.ts",
            "index-bench/LEVERS.md",
            "native-bin.ts",
            "runner.ts",
            "scenarios/conflict-synthesis.ts",
            "scenarios/docblocks-integration.ts",
            "scenarios/docblocks-theme-roundtrip.ts",
            "scenarios/first-projects.ts",
            "scenarios/helpers.ts",
            "scenarios/incident-postmortem.ts",
            "scenarios/index.ts",
            "scenarios/ops-runbook.ts",
            "scenarios/powerpoint-sources.ts",
            "scenarios/tool-provenance.ts",
            "scenarios/tool-routing-format.ts",
            "scenarios/work-routing.ts",
            "stale-step-session.ts",
            "success-check.ts",
            "suites.ts",
            "work-routing/analyze.ts",
            "work-routing/classify.ts",
            "work-routing/probes.ts",
            "work-routing/rescore.ts"
          ]
        },
        {
          "component": "runtime",
          "paths": [
            "@bendyline/gezel-mcp/package.json",
            "@bendyline/gezel-service/dependency/@bendyline/docblocks-cli",
            "@bendyline/gezel/dist/browser.js",
            "@bendyline/gezel/dist/chunk-3Z6PBVZM.js",
            "@bendyline/gezel/dist/chunk-5BXY37IX.js",
            "@bendyline/gezel/dist/chunk-6ASB7JVJ.js",
            "@bendyline/gezel/dist/chunk-6RY2SMVG.js",
            "@bendyline/gezel/dist/chunk-6YKM4W6M.js",
            "@bendyline/gezel/dist/chunk-7BFSCPZK.js",
            "@bendyline/gezel/dist/chunk-7QFCY42H.js",
            "@bendyline/gezel/dist/chunk-AOKPYNLV.js",
            "@bendyline/gezel/dist/chunk-B7GVWVXH.js",
            "@bendyline/gezel/dist/chunk-BB6SJOHT.js",
            "@bendyline/gezel/dist/chunk-CN3XHZA2.js",
            "@bendyline/gezel/dist/chunk-CZMCPJOM.js",
            "@bendyline/gezel/dist/chunk-DG6IPRC4.js",
            "@bendyline/gezel/dist/chunk-DUS3LUJM.js",
            "@bendyline/gezel/dist/chunk-HKEUN4VQ.js",
            "@bendyline/gezel/dist/chunk-KK2H55XN.js",
            "@bendyline/gezel/dist/chunk-KX3ESUXI.js",
            "@bendyline/gezel/dist/chunk-OY2IFYYQ.js",
            "@bendyline/gezel/dist/chunk-PF7H4CIR.js",
            "@bendyline/gezel/dist/chunk-PL6T3IO4.js",
            "@bendyline/gezel/dist/chunk-XJYBP6IZ.js",
            "@bendyline/gezel/dist/chunk-XTAG6E6T.js",
            "@bendyline/gezel/dist/chunk-YFNRCWMC.js",
            "@bendyline/gezel/dist/index.js",
            "@bendyline/gezel/dist/local-loop/index.js",
            "@bendyline/gezel/dist/markdown/index.js",
            "@bendyline/gezel/dist/native/index.js",
            "@bendyline/gezel/dist/paths.js",
            "@bendyline/gezel/dist/runtime/index.js",
            "@bendyline/gezel/dist/schemas/app-models.js",
            "@bendyline/gezel/dist/schemas/index.js",
            "@bendyline/gezel/dist/schemas/queue-status.js",
            "@bendyline/gezel/dist/speech-models/index.js"
          ]
        },
        {
          "component": "source",
          "paths": [
            "evals/src/auto-answer.ts",
            "evals/src/batch.ts",
            "evals/src/bin/ab-work-routing.ts",
            "evals/src/bin/image-bench.ts",
            "evals/src/craftbooks/adapter.ts",
            "evals/src/craftbooks/authoring/author-linear.ts",
            "evals/src/craftbooks/authoring/edit-midtask.ts",
            "evals/src/craftbooks/authoring/export-generalize.ts",
            "evals/src/craftbooks/authoring/gate-script.ts",
            "evals/src/craftbooks/authoring/helpers.ts",
            "evals/src/craftbooks/authoring/route-multi.ts",
            "evals/src/craftbooks/bundled-toolsets.ts",
            "evals/src/craftbooks/scenario.ts",
            "evals/src/craftbooks/specs.ts",
            "evals/src/craftbooks/test-spec-loader.ts",
            "evals/src/craftbooks/types.ts",
            "evals/src/docblocks-runtime.ts",
            "evals/src/engine-context.ts",
            "evals/src/fixtures/office-documents.ts",
            "evals/src/image-bench/IMAGE-BENCH.md",
            "evals/src/image-bench/alignment.ts",
            "evals/src/image-bench/judge.ts",
            "evals/src/image-bench/memory.ts",
            "evals/src/image-bench/prompts.ts",
            "evals/src/image-bench/sanity.ts",
            "evals/src/image-bench/sd-log.ts",
            "evals/src/index-bench/LEVERS.md",
            "evals/src/native-bin.ts",
            "evals/src/runner.ts",
            "evals/src/scenarios/conflict-synthesis.ts",
            "evals/src/scenarios/docblocks-integration.ts",
            "evals/src/scenarios/docblocks-theme-roundtrip.ts",
            "evals/src/scenarios/first-projects.ts",
            "evals/src/scenarios/helpers.ts",
            "evals/src/scenarios/incident-postmortem.ts",
            "evals/src/scenarios/index.ts",
            "evals/src/scenarios/ops-runbook.ts",
            "evals/src/scenarios/powerpoint-sources.ts",
            "evals/src/scenarios/tool-provenance.ts",
            "evals/src/scenarios/tool-routing-format.ts",
            "evals/src/scenarios/work-routing.ts",
            "evals/src/stale-step-session.ts",
            "evals/src/success-check.ts",
            "evals/src/suites.ts",
            "evals/src/work-routing/analyze.ts",
            "evals/src/work-routing/classify.ts",
            "evals/src/work-routing/probes.ts",
            "evals/src/work-routing/rescore.ts"
          ]
        }
      ],
      "observedSourceCommit": "f9a398ae1ee724aae45a55e62b796d3116fbdd8d",
      "recordedAt": "2026-10-22T05:44:23.941Z",
      "evidence": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/identity-drift-after-trial.json"
    },
    "performanceObservations": [
      {
        "firstObservedAt": "2026-10-20T00:00:55.687Z",
        "foreignEnginePid": 2134,
        "affectedCell": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/performance-overlap-2026-10-20T0000.json"
      },
      {
        "firstObservedDuringHeartbeat": "2026-10-21T14:11:38.955Z",
        "foreignEnginePid": 46551,
        "affectedCell": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/performance-overlap-2026-10-21T1411.json",
        "note": "Foreign DocBlocks Qwen engine overlapped TDD inference through at least 14:12:25Z and was absent at 14:16:58Z. Possible resource contention only; no causal attribution or outcome/score change."
      },
      {
        "firstObservedAt": "2026-10-21T14:26:53Z",
        "foreignEnginePid": 87609,
        "affectedCell": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/performance-overlap-2026-10-21T1426.json",
        "note": "Foreign Gemma engine active while campaign task is idle; no causal attribution or outcome change."
      },
      {
        "firstObservedAt": "2026-10-21T14:57:56Z",
        "foreignEnginePid": 43924,
        "affectedCell": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "evidence": "evals/runs/scorecard-seven-models-2026-10-15-proposal/performance-overlap-2026-10-21T1457.json",
        "note": "Foreign Qwen engine active during model copy; campaign inference has not started."
      }
    ],
    "scope": "Four complete models plus partial Ornith; Nemotron and MiniCPM unrun. No seven-model ranking is supported."
  },
  "trials": [
    {
      "trialId": "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 1
      },
      "startedAt": "2026-10-16T04:34:42.565Z",
      "finishedAt": "2026-10-16T04:35:31.961Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 49396,
        "timing": {
          "startedAt": "2026-10-16T04:34:42.565Z",
          "finishedAt": "2026-10-16T04:35:31.961Z",
          "timeToFirstArtifactMs": 45443,
          "timeToLastArtifactWriteMs": 45447,
          "timeToFirstTokenMs": 7635,
          "firstTurnTtftMs": 5786,
          "timeToFirstToolCallMs": 18788
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 170.47,
        "prefillTokensPerSec": 4500.21,
        "decodeTokens": 6095,
        "peakRssMb": 17338,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T04:37:16.132Z",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "trialId": "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/workspace/tic-tac-toe-game/index.html",
        "artifactSha256": "2559bb3e4bfacceb26312e1120fe13cacd06f33294748e6a03da9ebde6801a11",
        "artifactBytes": 2694,
        "artifactLimitBytes": 20480,
        "artifactClipped": false,
        "status": "judged",
        "rubric": "default-html",
        "scoreAxes": {
          "visualQuality": 4,
          "functionalCompleteness": 7,
          "codeQuality": 7,
          "polish": 3
        },
        "meanScore": 5.3,
        "justification": "Playable two-player game with win/draw handling, reset and visible turn status. The centered grid is legible but plain; initial cells show 0–8 while reset makes them blank. Fixed cell sizes, no win highlighting and little visual feedback limit polish.",
        "axisJustifications": {
          "visualQuality": "A clear centered grid and readable marks, but default gray buttons and a bare page feel unfinished.",
          "functionalCompleteness": "Alternating turns, all eight winning lines, draw detection and reset are implemented; runtime assertions passed. Initial numbered cells differ from the reset state.",
          "codeQuality": "Small readable functions and straightforward state handling; no external dependencies. Initial/reset cell content should be consistent.",
          "polish": "No win-line treatment, animation or tailored interaction states; fixed cell dimensions lack a responsive layout."
        },
        "evidenceMode": "Read final HTML source using the frozen rubric; deterministic runtime evidence considered separately from visual source review."
      }
    },
    {
      "trialId": "petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 1
      },
      "startedAt": "2026-10-16T04:36:08.035Z",
      "finishedAt": "2026-10-16T04:37:42.957Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 94922,
        "timing": {
          "startedAt": "2026-10-16T04:36:08.035Z",
          "finishedAt": "2026-10-16T04:37:42.957Z",
          "timeToFirstArtifactMs": 90940,
          "timeToLastArtifactWriteMs": 90945,
          "timeToFirstTokenMs": 7451,
          "firstTurnTtftMs": 5802,
          "timeToFirstToolCallMs": 15503
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 154.95,
        "prefillTokensPerSec": 2769.34,
        "decodeTokens": 9984,
        "peakRssMb": 25658,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.581Z",
        "trialId": "petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/workspace/pet-shop-website/index.html",
        "artifactBytes": 2357,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 5,
          "functionalCompleteness": 5,
          "codeQuality": 5,
          "polish": 3
        },
        "meanScore": 4.5,
        "justification": "Legible generic shop with the requested logo reference. Card layout styles are never applied to the container, and Buy All repeatedly raises prices and accumulates totals. The added purchase interaction is misleading; branding and interaction polish are minimal.",
        "artifactSha256": "b64dbee364dfd2f09041aaa537ed69fcbcb1149ed1f6ccec56511bbe44dc81c9",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 1
      },
      "startedAt": "2026-10-16T04:38:18.977Z",
      "finishedAt": "2026-10-16T04:47:44.077Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 565100,
        "timing": {
          "startedAt": "2026-10-16T04:38:18.977Z",
          "finishedAt": "2026-10-16T04:47:44.077Z",
          "timeToFirstArtifactMs": 164546,
          "timeToLastArtifactWriteMs": 557163,
          "timeToFirstTokenMs": 7408,
          "firstTurnTtftMs": 5868,
          "timeToFirstToolCallMs": 33146
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 117.37,
        "prefillTokensPerSec": 2298.1,
        "decodeTokens": 53385,
        "peakRssMb": 19040,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.586Z",
        "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 5725,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 4,
          "functionalCompleteness": 1,
          "codeQuality": 2,
          "polish": 2
        },
        "meanScore": 2.3,
        "justification": "Movement and a HUD exist, but combat is broken: player bullets never hit enemies, both collision checks use enemy bullets, the fire timer mixes seconds with 1200, and score never changes. Minimal rectangles, no restart or instructions. Automated gate success does not establish playable combat.",
        "artifactSha256": "662c6e6bab5d037ea817e0aabd74030ce16e09d3001d3f81da89cd16cd49ade4",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 1
      },
      "startedAt": "2026-10-16T04:48:21.034Z",
      "finishedAt": "2026-10-16T04:50:45.322Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 144288,
        "timing": {
          "startedAt": "2026-10-16T04:48:21.034Z",
          "finishedAt": "2026-10-16T04:50:45.322Z",
          "timeToFirstArtifactMs": 18106,
          "timeToLastArtifactWriteMs": 140301,
          "timeToFirstTokenMs": 6620,
          "firstTurnTtftMs": 5015,
          "timeToFirstToolCallMs": 18110
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 137.27,
        "prefillTokensPerSec": 2851.73,
        "decodeTokens": 13762,
        "peakRssMb": 18193,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.588Z",
        "trialId": "schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 1
      },
      "startedAt": "2026-10-16T04:50:55.139Z",
      "finishedAt": "2026-10-16T05:04:42.745Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"failing-tests-spec:3:targetnone:frcekrw7:rp0:rf0\" stuck for 8m — 7 re-writes (29 tool calls) without sniff movement. Artifact produced but never reached success (retry-loop nudge was sent and ignored).",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 827606,
        "timing": {
          "startedAt": "2026-10-16T04:50:55.139Z",
          "finishedAt": "2026-10-16T05:04:42.745Z",
          "timeToFirstArtifactMs": 21561,
          "timeToLastArtifactWriteMs": 820395,
          "timeToFirstTokenMs": 7200,
          "firstTurnTtftMs": 5573,
          "timeToFirstToolCallMs": 7633
        },
        "toolCalls": 66,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 113.52,
        "prefillTokensPerSec": 2371.85,
        "decodeTokens": 58589,
        "peakRssMb": 25036,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.588Z",
        "trialId": "failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:04:54.300Z",
      "finishedAt": "2026-10-16T05:05:56.961Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 62661,
        "timing": {
          "startedAt": "2026-10-16T05:04:54.300Z",
          "finishedAt": "2026-10-16T05:05:56.961Z",
          "timeToFirstArtifactMs": 58990,
          "timeToLastArtifactWriteMs": 58990,
          "timeToFirstTokenMs": 7263,
          "firstTurnTtftMs": 5626,
          "timeToFirstToolCallMs": 8102
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 140.7,
        "prefillTokensPerSec": 3034.52,
        "decodeTokens": 7110,
        "peakRssMb": 16793,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.589Z",
        "trialId": "symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:06:06.733Z",
      "finishedAt": "2026-10-16T05:11:58.023Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 351290,
        "timing": {
          "startedAt": "2026-10-16T05:06:06.733Z",
          "finishedAt": "2026-10-16T05:11:58.023Z",
          "timeToFirstArtifactMs": 46394,
          "timeToLastArtifactWriteMs": 344806,
          "timeToFirstTokenMs": 7467,
          "firstTurnTtftMs": 5881,
          "timeToFirstToolCallMs": 8458
        },
        "toolCalls": 34,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 112.4,
        "prefillTokensPerSec": 2439.23,
        "decodeTokens": 35472,
        "peakRssMb": 17268,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.589Z",
        "trialId": "data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:12:07.451Z",
      "finishedAt": "2026-10-16T05:14:41.203Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 153752,
        "timing": {
          "startedAt": "2026-10-16T05:12:07.451Z",
          "finishedAt": "2026-10-16T05:14:41.203Z",
          "timeToFirstArtifactMs": 68180,
          "timeToLastArtifactWriteMs": 150171,
          "timeToFirstTokenMs": 8645,
          "firstTurnTtftMs": 6934,
          "timeToFirstToolCallMs": 9947
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 105.24,
        "prefillTokensPerSec": 3537.54,
        "decodeTokens": 13790,
        "peakRssMb": 16993,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.590Z",
        "trialId": "incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:14:50.100Z",
      "finishedAt": "2026-10-16T05:15:27.450Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37350,
        "timing": {
          "startedAt": "2026-10-16T05:14:50.100Z",
          "finishedAt": "2026-10-16T05:15:27.450Z",
          "timeToFirstArtifactMs": 30029,
          "timeToLastArtifactWriteMs": 33172,
          "timeToFirstTokenMs": 10633,
          "firstTurnTtftMs": 8934,
          "timeToFirstToolCallMs": 11387
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 118.41,
        "prefillTokensPerSec": 2526.5,
        "decodeTokens": 2544,
        "peakRssMb": 16808,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.590Z",
        "trialId": "ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:15:36.306Z",
      "finishedAt": "2026-10-16T05:16:39.049Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 62743,
        "timing": {
          "startedAt": "2026-10-16T05:15:36.306Z",
          "finishedAt": "2026-10-16T05:16:39.049Z",
          "timeToFirstArtifactMs": 20802,
          "timeToLastArtifactWriteMs": 57497,
          "timeToFirstTokenMs": 7526,
          "firstTurnTtftMs": 5953,
          "timeToFirstToolCallMs": 10539
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 134.39,
        "prefillTokensPerSec": 2359.36,
        "decodeTokens": 6902,
        "peakRssMb": 16651,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.590Z",
        "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:16:48.727Z",
      "finishedAt": "2026-10-16T05:22:51.629Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/9a3c1936 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 76.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 362902,
        "timing": {
          "startedAt": "2026-10-16T05:16:48.727Z",
          "finishedAt": "2026-10-16T05:22:51.629Z",
          "timeToFirstArtifactMs": 34735,
          "timeToLastArtifactWriteMs": 288350,
          "timeToFirstTokenMs": 10144,
          "firstTurnTtftMs": 8555,
          "timeToFirstToolCallMs": 12652
        },
        "toolCalls": 40,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 101.11,
        "prefillTokensPerSec": 2170.93,
        "decodeTokens": 29798,
        "peakRssMb": 18715,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.590Z",
        "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb",
        "cellId": "r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:23:09.503Z",
      "finishedAt": "2026-10-16T05:24:18.015Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 68512,
        "timing": {
          "startedAt": "2026-10-16T05:23:09.503Z",
          "finishedAt": "2026-10-16T05:24:18.015Z",
          "timeToFirstArtifactMs": 44992,
          "timeToLastArtifactWriteMs": 67186,
          "timeToFirstTokenMs": 8025,
          "firstTurnTtftMs": 5872,
          "timeToFirstToolCallMs": 9974
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 108.23,
        "prefillTokensPerSec": 2297.81,
        "decodeTokens": 5437,
        "peakRssMb": 16840,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.591Z",
        "trialId": "constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:24:35.040Z",
      "finishedAt": "2026-10-16T05:27:26.360Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 171320,
        "timing": {
          "startedAt": "2026-10-16T05:24:35.040Z",
          "finishedAt": "2026-10-16T05:27:26.360Z",
          "timeToFirstArtifactMs": 67160,
          "timeToLastArtifactWriteMs": 169123,
          "timeToFirstTokenMs": 8936,
          "firstTurnTtftMs": 6000,
          "timeToFirstToolCallMs": 10634
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 98.71,
        "prefillTokensPerSec": 3223.12,
        "decodeTokens": 15423,
        "peakRssMb": 16932,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.591Z",
        "trialId": "records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:27:48.225Z",
      "finishedAt": "2026-10-16T05:29:17.030Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 88805,
        "timing": {
          "startedAt": "2026-10-16T05:27:48.225Z",
          "finishedAt": "2026-10-16T05:29:17.030Z",
          "timeToFirstArtifactMs": 84862,
          "timeToLastArtifactWriteMs": 84862,
          "timeToFirstTokenMs": 9909,
          "firstTurnTtftMs": 7624,
          "timeToFirstToolCallMs": 50625
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 101.58,
        "prefillTokensPerSec": 1827.13,
        "decodeTokens": 7659,
        "peakRssMb": 16778,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.592Z",
        "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:29:35.433Z",
      "finishedAt": "2026-10-16T05:32:49.618Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 194185,
        "timing": {
          "startedAt": "2026-10-16T05:29:35.433Z",
          "finishedAt": "2026-10-16T05:32:49.618Z",
          "timeToFirstArtifactMs": 33234,
          "timeToLastArtifactWriteMs": 191398,
          "timeToFirstTokenMs": 11514,
          "firstTurnTtftMs": 9253,
          "timeToFirstToolCallMs": 14911
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 97.71,
        "prefillTokensPerSec": 1875.74,
        "decodeTokens": 6698,
        "peakRssMb": 17049,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.592Z",
        "trialId": "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1969,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 7,
          "grounding": 6,
          "actionability": 7,
          "brevity": 7
        },
        "meanScore": 6.8,
        "justification": "Clear decisions, named owners, dates, dependency and unresolved risks match most of the transcript. It blurs scheduling the annual-plan review by August 30 with holding the review on that date, and presents that action as a decision. Otherwise concise and directly usable.",
        "artifactSha256": "afbce6ae53597ed56d489b562eab754246af370636fe34d96a37bcb5aae73aca",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 1
      },
      "startedAt": "2026-10-16T05:33:06.868Z",
      "finishedAt": "2026-10-16T06:02:05.862Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: ivana/a97b834e exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model couldn't apply its file edit after 5 tries, so the turn was stopped. Try sending your message again, or switch to a more capable model in Settings.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.9,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 66.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1738994,
        "timing": {
          "startedAt": "2026-10-16T05:33:06.868Z",
          "finishedAt": "2026-10-16T06:02:05.862Z",
          "timeToFirstArtifactMs": 50640,
          "timeToLastArtifactWriteMs": 1012086,
          "timeToFirstTokenMs": 16029,
          "firstTurnTtftMs": 8758,
          "timeToFirstToolCallMs": 21432
        },
        "toolCalls": 64,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.16,
        "prefillTokensPerSec": 1789.75,
        "decodeTokens": 140184,
        "peakRssMb": 21197,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.594Z",
        "trialId": "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 4648,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 2
        },
        "meanScore": 3.5,
        "justification": "The scope separates purpose, assumptions, formulas and acceptance criteria, but its financial contract is wrong: stated defaults imply annual net profit of 728,400, not 70,800. Multiplying net profit by 30 does not calculate cash runway, and churn is listed without calculation logic.",
        "artifactSha256": "4b2809e10f9b93ea71b8ec7b6f27c9578e8cd9272c06a8e9c54f978867f966eb",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-16T06:02:32.538Z",
      "finishedAt": "2026-10-16T06:07:59.772Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/86d66726 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 84.6% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 327234,
        "timing": {
          "startedAt": "2026-10-16T06:02:32.538Z",
          "finishedAt": "2026-10-16T06:07:59.772Z",
          "timeToFirstArtifactMs": 49614,
          "timeToLastArtifactWriteMs": 255347,
          "timeToFirstTokenMs": 14230,
          "firstTurnTtftMs": 11497,
          "timeToFirstToolCallMs": 18012
        },
        "toolCalls": 34,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.7,
        "prefillTokensPerSec": 1739.02,
        "decodeTokens": 22868,
        "peakRssMb": 18891,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.596Z",
        "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 1
      },
      "startedAt": "2026-10-16T06:08:21.746Z",
      "finishedAt": "2026-10-16T06:09:55.954Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 94208,
        "timing": {
          "startedAt": "2026-10-16T06:08:21.746Z",
          "finishedAt": "2026-10-16T06:09:55.954Z",
          "timeToFirstArtifactMs": 91626,
          "timeToLastArtifactWriteMs": 91626,
          "timeToFirstTokenMs": 11047,
          "firstTurnTtftMs": 8466,
          "timeToFirstToolCallMs": 19604
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 98.21,
        "prefillTokensPerSec": 1869.25,
        "decodeTokens": 7407,
        "peakRssMb": 17176,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.596Z",
        "trialId": "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1103,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 3,
          "candor": 7,
          "usefulness": 6
        },
        "meanScore": 5.3,
        "justification": "The known theme values and page counts are reproduced, and Callout Emphasis is plainly unresolved. However, it invents a successful transfer inventory of Title, Heading 1/2, Body and Table Grid beyond the supplied tool facts. Adding a matching brand style is an actionable recommendation.",
        "artifactSha256": "85e7bdffd6949434408031d05b3296cd6ca57e6e157ff94b3a5efb73e1979efd",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 1
      },
      "startedAt": "2026-10-16T06:10:16.467Z",
      "finishedAt": "2026-10-16T06:19:31.406Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 554939,
        "timing": {
          "startedAt": "2026-10-16T06:10:16.467Z",
          "finishedAt": "2026-10-16T06:19:31.406Z",
          "timeToFirstArtifactMs": 72209,
          "timeToLastArtifactWriteMs": 552203,
          "timeToFirstTokenMs": 18467,
          "firstTurnTtftMs": 11264,
          "timeToFirstToolCallMs": 20512
        },
        "toolCalls": 73,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 94.82,
        "prefillTokensPerSec": 1895.87,
        "decodeTokens": 43426,
        "peakRssMb": 19737,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.599Z",
        "trialId": "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1400,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 7,
          "outline-parity": 8,
          "grounding": 8,
          "convertibility": 7
        },
        "meanScore": 7.5,
        "justification": "The source is cited throughout; all seven locked outline titles appear once and in order. Metrics and actions match the supplied brief. Markdown is conversion-ready and concise, although the SKU table is a placeholder and the evidence labels are repetitive.",
        "artifactSha256": "4326106e848c74f74cdd2929440c33a1ba46b67806b1251186d7eb42dcc23138",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 1
      },
      "startedAt": "2026-10-16T06:19:54.490Z",
      "finishedAt": "2026-10-16T06:26:07.381Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: claim-calibration. Last failure: every mention of the “first programmer” label must be hedged in the same sentence as debated shorthand",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 372891,
        "timing": {
          "startedAt": "2026-10-16T06:19:54.490Z",
          "finishedAt": "2026-10-16T06:26:07.381Z",
          "timeToFirstArtifactMs": 158802,
          "timeToLastArtifactWriteMs": 369337,
          "timeToFirstTokenMs": 8243,
          "firstTurnTtftMs": 6215,
          "timeToFirstToolCallMs": 30899
        },
        "toolCalls": 15,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 88.91,
        "prefillTokensPerSec": 1858.1,
        "decodeTokens": 27904,
        "peakRssMb": 18130,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.601Z",
        "trialId": "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5585,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 2,
          "citationQuality": 4,
          "calibration": 4
        },
        "meanScore": 3.5,
        "justification": "The brief preserves the basic 1842/1843 chronology, but adds unsupported quotations, publication and abandonment details, and a claim that notes survive only in posthumous copies. Frequent source tags conceal those gaps. First-programmer caveats coexist with unqualified first-program claims.",
        "artifactSha256": "c0074aa463385943b0a3a0e4003712b412380cfec2f0dc1c7f3b2604f2b4cdbf",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 1
      },
      "startedAt": "2026-10-16T06:26:29.061Z",
      "finishedAt": "2026-10-16T07:56:35.776Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "hit hard ceiling (2700000ms + 45m progress extensions) — last hard progress 31s before the end; the deliverable never closed",
        "failureMode": "timeout",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 1.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 5406715,
        "timing": {
          "startedAt": "2026-10-16T06:26:29.061Z",
          "finishedAt": "2026-10-16T07:56:35.776Z",
          "timeToFirstArtifactMs": 619139,
          "timeToLastArtifactWriteMs": 5253063,
          "timeToFirstTokenMs": 21061,
          "firstTurnTtftMs": 13565,
          "timeToFirstToolCallMs": 26130
        },
        "toolCalls": 488,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 93.74,
        "prefillTokensPerSec": 2044.97,
        "decodeTokens": 434691,
        "peakRssMb": 20717,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.604Z",
        "trialId": "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/artifacts/calendar-desk-week-plan-eval/tasks/11/report.md",
        "artifactBytes": 2578,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 0,
          "prioritization": 1,
          "conflict-handling": 0,
          "realism": 1
        },
        "meanScore": 0.5,
        "justification": "The scenario brief requires eight calendar events and a Tuesday overlap; this plan declares every day empty and schedules only follow-ups. Its header calls October 16 Monday while listing it as Friday below. Focus allocations and conflict claims are therefore unsupported. Calendar query/date provenance needs a separate grader audit.",
        "artifactSha256": "74f58d4c0a201b41237b10cce3bd7f2b9b6d18b167b586e6b4100f4811551167",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 1
      },
      "startedAt": "2026-10-16T07:56:50.607Z",
      "finishedAt": "2026-10-16T08:08:30.513Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 699906,
        "timing": {
          "startedAt": "2026-10-16T07:56:50.607Z",
          "finishedAt": "2026-10-16T08:08:30.513Z",
          "timeToFirstArtifactMs": 69327,
          "timeToLastArtifactWriteMs": 694742,
          "timeToFirstTokenMs": 16618,
          "firstTurnTtftMs": 9954,
          "timeToFirstToolCallMs": 20346
        },
        "toolCalls": 59,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 108.04,
        "prefillTokensPerSec": 2466.64,
        "decodeTokens": 62222,
        "peakRssMb": 22165,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.605Z",
        "trialId": "craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2062,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared axes safety/mapping/signals/containment describe mock-service deployment, while the requested artifact is a statistical A/B readout. No meaningful quality score is assigned until that evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:08:40.566Z",
      "finishedAt": "2026-10-16T08:12:41.526Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 240960,
        "timing": {
          "startedAt": "2026-10-16T08:08:40.566Z",
          "finishedAt": "2026-10-16T08:12:41.526Z",
          "timeToFirstArtifactMs": 38574,
          "timeToLastArtifactWriteMs": 188619,
          "timeToFirstTokenMs": 14400,
          "firstTurnTtftMs": 7803,
          "timeToFirstToolCallMs": 19344
        },
        "toolCalls": 26,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 107.34,
        "prefillTokensPerSec": 2617.16,
        "decodeTokens": 20960,
        "peakRssMb": 19072,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.606Z",
        "trialId": "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 4741,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 7,
          "grounding": 3,
          "structure": 7,
          "actionability": 6
        },
        "meanScore": 5.8,
        "justification": "The six seeded sources and their numerical findings are retained in consistent three-part annotations. However, Allen acquires invented volume/issue/pages, Garcia an unsupported single-region limitation, and several sources unsupported peer-review or authority claims. Relevance explanations are useful but certainty is overstated.",
        "artifactSha256": "b6e63689d9b6132fff32802044d22ca5daac83bbce138361b4a020b70bac6df6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:12:50.708Z",
      "finishedAt": "2026-10-16T08:18:01.318Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 310610,
        "timing": {
          "startedAt": "2026-10-16T08:12:50.708Z",
          "finishedAt": "2026-10-16T08:18:01.318Z",
          "timeToFirstArtifactMs": 61553,
          "timeToLastArtifactWriteMs": 230408,
          "timeToFirstTokenMs": 14271,
          "firstTurnTtftMs": 7696,
          "timeToFirstToolCallMs": 22636
        },
        "toolCalls": 39,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 108.73,
        "prefillTokensPerSec": 2634.94,
        "decodeTokens": 27752,
        "peakRssMb": 19149,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.608Z",
        "trialId": "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh",
        "cellId": "r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1741,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 7,
          "structure": 7,
          "convertibility": 8
        },
        "meanScore": 7.3,
        "justification": "The supplied pilot metrics, unresolved complaint and three next actions survive in a clear summary, table and findings structure. Markdown is complete for DOCX conversion. Some interpretation and expanded action scope go beyond the sparse brief, but no new metric is introduced.",
        "artifactSha256": "5659fe9555255a1f09263fc09d2847eef729d965a6af2a8a985a6b1c540b1fb7",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:18:10.226Z",
      "finishedAt": "2026-10-16T08:21:29.346Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 199120,
        "timing": {
          "startedAt": "2026-10-16T08:18:10.226Z",
          "finishedAt": "2026-10-16T08:21:29.346Z",
          "timeToFirstArtifactMs": 57789,
          "timeToLastArtifactWriteMs": 193853,
          "timeToFirstTokenMs": 7768,
          "firstTurnTtftMs": 6051,
          "timeToFirstToolCallMs": 8518
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 110.13,
        "prefillTokensPerSec": 2426.21,
        "decodeTokens": 17051,
        "peakRssMb": 18578,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.610Z",
        "trialId": "fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:21:38.262Z",
      "finishedAt": "2026-10-16T08:22:16.425Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 38163,
        "timing": {
          "startedAt": "2026-10-16T08:21:38.262Z",
          "finishedAt": "2026-10-16T08:22:16.425Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 11204,
          "firstTurnTtftMs": 9623,
          "timeToFirstToolCallMs": 16045
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 113.8,
        "prefillTokensPerSec": 2394.62,
        "decodeTokens": 2352,
        "peakRssMb": 16849,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.610Z",
        "trialId": "dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:22:25.596Z",
      "finishedAt": "2026-10-16T08:27:36.476Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 310880,
        "timing": {
          "startedAt": "2026-10-16T08:22:25.596Z",
          "finishedAt": "2026-10-16T08:27:36.476Z",
          "timeToFirstArtifactMs": 290052,
          "timeToLastArtifactWriteMs": 304792,
          "timeToFirstTokenMs": 8923,
          "firstTurnTtftMs": 7267,
          "timeToFirstToolCallMs": 44696
        },
        "toolCalls": 26,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 114.09,
        "prefillTokensPerSec": 2342.79,
        "decodeTokens": 29268,
        "peakRssMb": 17741,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.611Z",
        "trialId": "interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:27:46.214Z",
      "finishedAt": "2026-10-16T08:48:20.332Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 724s — model busy but not delivering (hard digest stuck at a9131b9d202f08e2)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_NO_PRODUCED_OUTPUT",
            "summary": "No authored artifact event or scored sniff output was observed."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1234118,
        "timing": {
          "startedAt": "2026-10-16T08:27:46.214Z",
          "finishedAt": "2026-10-16T08:48:20.332Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 13904,
          "firstTurnTtftMs": 7342,
          "timeToFirstToolCallMs": 15289
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 103.99,
        "prefillTokensPerSec": 2170.72,
        "decodeTokens": 46879,
        "peakRssMb": 18280,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.611Z",
        "trialId": "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "noArtifact",
        "comparisonEligible": false,
        "justification": "The exact artifact basename declared by this scenario was not found in the captured workspace or artifacts drawer."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 1
      },
      "startedAt": "2026-10-16T08:48:29.350Z",
      "finishedAt": "2026-10-16T09:04:42.678Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 973328,
        "timing": {
          "startedAt": "2026-10-16T08:48:29.350Z",
          "finishedAt": "2026-10-16T09:04:42.678Z",
          "timeToFirstArtifactMs": 174083,
          "timeToLastArtifactWriteMs": 890226,
          "timeToFirstTokenMs": 15687,
          "firstTurnTtftMs": 9157,
          "timeToFirstToolCallMs": 22523
        },
        "toolCalls": 42,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 102.1,
        "prefillTokensPerSec": 2560.43,
        "decodeTokens": 81306,
        "peakRssMb": 20267,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.611Z",
        "trialId": "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 9377,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 2
        },
        "meanScore": 3.5,
        "justification": "A staged review and corrected snippets exist, but the report contradicts its own finding counts and compatibility verdict. It falsely flags optional nullable email, invents schema facts, puts Idempotency-Key under invalid operation-level headers, and proposes nonstandard version/deprecation fields as fixes.",
        "artifactSha256": "cf009f73f3c8e73019dee8382642307f3baea931e9a4973da62e3c06f0bbf125",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 1
      },
      "startedAt": "2026-10-16T09:04:52.359Z",
      "finishedAt": "2026-10-16T09:07:01.447Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 129088,
        "timing": {
          "startedAt": "2026-10-16T09:04:52.359Z",
          "finishedAt": "2026-10-16T09:07:01.447Z",
          "timeToFirstArtifactMs": 35157,
          "timeToLastArtifactWriteMs": 123579,
          "timeToFirstTokenMs": 7924,
          "firstTurnTtftMs": 6300,
          "timeToFirstToolCallMs": 8839
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 144.13,
        "prefillTokensPerSec": 3294.79,
        "decodeTokens": 13333,
        "peakRssMb": 17946,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.614Z",
        "trialId": "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 4156,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 7,
          "systemic analysis": 4,
          "remediation": 6
        },
        "meanScore": 5.7,
        "justification": "All three real admin vulnerabilities have concrete source-to-sink evidence, and safe files are identified correctly. The root-cause account speculates about authors and contradicts itself on the safe allow-list. JSON includes a concrete parameterized SQL fix, but the report omits it and shell remediation remains vague.",
        "artifactSha256": "a6cc68a9475e43586e6bdc4c19d6dbda5822ddc5ac600e3c1984d727e53d9ec6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 1
      },
      "startedAt": "2026-10-16T09:07:11.869Z",
      "finishedAt": "2026-10-16T09:22:44.281Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"large-pr-review:4:targetpr-review.md:fr1i0sit8:rp0:rf0\" stuck for 8m — 7 re-writes (9 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_UNGRADED_AUTHORED_OUTPUT",
            "summary": "An authored output exists without a sniff verdict; quality is capped at the rubric midpoint."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 932412,
        "timing": {
          "startedAt": "2026-10-16T09:07:11.869Z",
          "finishedAt": "2026-10-16T09:22:44.281Z",
          "timeToFirstArtifactMs": 235099,
          "timeToLastArtifactWriteMs": 882472,
          "timeToFirstTokenMs": 11896,
          "firstTurnTtftMs": 10042,
          "timeToFirstToolCallMs": 14972
        },
        "toolCalls": 39,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 72.61,
        "prefillTokensPerSec": 1573.31,
        "decodeTokens": 60821,
        "peakRssMb": 19742,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.615Z",
        "trialId": "large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 1
      },
      "startedAt": "2026-10-16T09:22:53.308Z",
      "finishedAt": "2026-10-16T09:28:17.692Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 324384,
        "timing": {
          "startedAt": "2026-10-16T09:22:53.308Z",
          "finishedAt": "2026-10-16T09:28:17.692Z",
          "timeToFirstArtifactMs": 62187,
          "timeToLastArtifactWriteMs": 291035,
          "timeToFirstTokenMs": 14822,
          "firstTurnTtftMs": 8290,
          "timeToFirstToolCallMs": 16074
        },
        "toolCalls": 84,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 113.32,
        "prefillTokensPerSec": 2841.6,
        "decodeTokens": 29282,
        "peakRssMb": 19378,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.615Z",
        "trialId": "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 1670,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 6,
          "Invariant discipline": 7,
          "Reconciled verification": 6
        },
        "meanScore": 6.3,
        "justification": "The residual scan, intentional changelog exclusion and three green tests are reported. Counts are presented consistently, but the initial site list omits a test import that the verification later claims was changed, and the summary does not clearly reconcile the log-string occurrence. Mechanical behavior checks passed.",
        "artifactSha256": "2ebd84e0b41eb754fbf61e871f17491fa0ae717af67834b4b5d46eea90ab9cc9",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 1
      },
      "startedAt": "2026-10-16T09:28:26.981Z",
      "finishedAt": "2026-10-16T09:50:42.689Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on tasks/1/baseline.md; missing signals stayed unchanged: tests/verify-refactor.mjs did not pass when run with node: exit=1\nCommand failed: /usr/local/bin/node tests/verify-refactor.mjs\nnode:internal/modules/run_main:107\n    triggerUncaughtException(\n    ^\n\nAssertionError [ERR_ASSERTION]: the title-case implementation appears 2 times across src/ - the duplication was not extracted to one shared place\n    at <sandbox>/tests/verify-refactor.mjs:47:8\n    at ModuleJob.run (node:internal/modules/esm/module_job:439:25)\n    at async node:internal/modules/esm/loader:643:26\n    at async asyncRunEntryPointWithESMLoader (node:internal/modules/run_main:101:5) {\n  generatedMessage: false,\n  code: 'ERR_ASSERTION',\n  actual: false,\n  expected: true,\n  operator: '==',\n  diff: 'simple'\n}\n\nNode.js v24.18.0\nREFACTOR_MODULE_ORACLE behavior ok, seeded workspace input(s) have not been read yet: source/refactor-brief.md, src/format-name.js contains forbidden content: format-name.js no longer carries its own title-case copy (matched \"toUpperCase\"), task-notes.md is missing required content: DONE note names the real test command — nothing in its 3850 bytes matches /\\bDONE\\b[\\s\\S]*(npm\\s+run\\s+test|node\\s+--test)/. Add that content.. Last failure: tests/verify-refactor.mjs did not pass when run with node: exit=1 Command failed: /usr/local/bin/node tests/verify-refactor.mjs node:internal/modules/run_main:107 triggerUncaughtException( ^ AssertionError [ERR_ASSERTION]: the title-case implementation appears 2 times across src/ - the duplication was not extracted to one shared place at <sandbox>/tests/verify-refactor.mjs:47:8 at ModuleJob.run (node:internal/modules/esm/module_job:439:25) at async node:internal/modules/esm/loader:643:26 at async asyncRunEntryPointWithESMLoader (node:internal/modules/run_main:101:5) { generatedMessage: false, code: 'ERR_ASSERTION', actual: false, expected: true, operator: '==', diff: 'simple' } Node.js v24.1",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 86.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1335708,
        "timing": {
          "startedAt": "2026-10-16T09:28:26.981Z",
          "finishedAt": "2026-10-16T09:50:42.689Z",
          "timeToFirstArtifactMs": 260157,
          "timeToLastArtifactWriteMs": 1331495,
          "timeToFirstTokenMs": 13138,
          "firstTurnTtftMs": 6578,
          "timeToFirstToolCallMs": 20215
        },
        "toolCalls": 154,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 109.66,
        "prefillTokensPerSec": 2443.86,
        "decodeTokens": 129299,
        "peakRssMb": 23525,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.618Z",
        "trialId": "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 1140,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 4,
          "Reversible-stage execution": 2,
          "Genuine deduplication": 0
        },
        "meanScore": 2,
        "justification": "The baseline describes tests and duplication, but stage notes are not faithful to the saved source: both format-name.js and normalize-name.js still implement normalization, report.js imports the former, and titleCaseName is no longer exported. Claimed successful extraction and no deviations contradict the actual files.",
        "artifactSha256": "700224d6384c5637557d875d78cc5b4cdcf36eb79f60cd625c33373fadd02c69",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 1
      },
      "startedAt": "2026-10-16T09:50:52.327Z",
      "finishedAt": "2026-10-16T10:19:26.634Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on workspace output; missing signals stayed unchanged: task sourced from craftbook bug-fix-tdd has not reached a terminal step, task-notes.md is missing required content: DONE note names the real test command — nothing in its 7577 bytes matches /\\bDONE\\b[\\s\\S]*(npm\\s+run\\s+test|node\\s+--test)/. Add that content.. Last failure: task sourced from craftbook bug-fix-tdd has not reached a terminal step",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1714307,
        "timing": {
          "startedAt": "2026-10-16T09:50:52.327Z",
          "finishedAt": "2026-10-16T10:19:26.634Z",
          "timeToFirstArtifactMs": 47422,
          "timeToLastArtifactWriteMs": 1710804,
          "timeToFirstTokenMs": 14849,
          "firstTurnTtftMs": 8283,
          "timeToFirstToolCallMs": 16311
        },
        "toolCalls": 144,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 89.46,
        "prefillTokensPerSec": 1777.15,
        "decodeTokens": 129525,
        "peakRssMb": 19259,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.620Z",
        "trialId": "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b",
        "cellId": "r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1773,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 6,
          "Fix minimality": 8,
          "Honest verification": 6
        },
        "meanScore": 6.7,
        "justification": "The saved fix is the minimal > to >= change, with a real regression test at the threshold. Companion repro/validation notes quote red and green runs. However, the final verification overclaims that every deterministic check passed although the workflow remained nonterminal, and review completion is not distinguished.",
        "artifactSha256": "f34d8bd93b57a68f124be16f6277560b7ffc67e76a038ea3f2f7b95c24823a49",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 1
      },
      "startedAt": "2026-10-16T10:19:36.090Z",
      "finishedAt": "2026-10-16T10:20:07.855Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 31765,
        "timing": {
          "startedAt": "2026-10-16T10:19:36.090Z",
          "finishedAt": "2026-10-16T10:20:07.855Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 11335,
          "firstTurnTtftMs": 9759,
          "timeToFirstToolCallMs": 17267
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 114.68,
        "prefillTokensPerSec": 2352.9,
        "decodeTokens": 1497,
        "peakRssMb": 16845,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.623Z",
        "trialId": "craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "A small HTML page exists, but this recipe-selection task declares no quality axes. Scoring that page with the game/site rubric would not grade the requested workflow."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 1
      },
      "startedAt": "2026-10-16T10:20:17.233Z",
      "finishedAt": "2026-10-16T10:38:22.753Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-edit-midtask:1:targetnone:fr16amiu3:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 80 tool calls (31 re-writes). Artifact produced but never reached success (retry-loop nudge was sent and ignored).",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 2.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 25% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1085520,
        "timing": {
          "startedAt": "2026-10-16T10:20:17.233Z",
          "finishedAt": "2026-10-16T10:38:22.753Z",
          "timeToFirstArtifactMs": 247791,
          "timeToLastArtifactWriteMs": 928277,
          "timeToFirstTokenMs": 13433,
          "firstTurnTtftMs": 11765,
          "timeToFirstToolCallMs": 16631
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.16,
        "prefillTokensPerSec": 1842.49,
        "decodeTokens": 77336,
        "peakRssMb": 18257,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.624Z",
        "trialId": "craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 1
      },
      "startedAt": "2026-10-16T10:38:31.750Z",
      "finishedAt": "2026-10-16T10:43:55.957Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 324207,
        "timing": {
          "startedAt": "2026-10-16T10:38:31.750Z",
          "finishedAt": "2026-10-16T10:43:55.957Z",
          "timeToFirstArtifactMs": 39958,
          "timeToLastArtifactWriteMs": 316201,
          "timeToFirstTokenMs": 12670,
          "firstTurnTtftMs": 6128,
          "timeToFirstToolCallMs": 13866
        },
        "toolCalls": 53,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 123.78,
        "prefillTokensPerSec": 2789.64,
        "decodeTokens": 32172,
        "peakRssMb": 19941,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.624Z",
        "trialId": "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 1563,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 3,
          "numbering continuity": 8,
          "professionalism": 5,
          "completeness": 5,
          "print-polish": 6
        },
        "meanScore": 5.4,
        "justification": "Totals and invoice numbers 042–044 reconcile to the ledger, and all three clients receive print-styled invoices. Kestrel is split into invented named SKUs priced 900/800/900 without ledger support. The original ledger still marks billable entries uninvoiced; payment instructions and consistent sender details are thin.",
        "artifactSha256": "2167896e9c0d0c75ecf5f2f82eba265443242986d2af986dcafab81dc85efb23",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 1
      },
      "startedAt": "2026-10-16T10:44:05.834Z",
      "finishedAt": "2026-10-16T10:53:52.606Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on craftbook: CSV Order Cleanup; missing signals stayed unchanged: task order-intake-cleanup/1 (from craftbook \"csv-order-cleanup\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count., notes/anomalies.md is 0 bytes, need ≥ 80, out/report.md is 0 bytes, need ≥ 80, out/orders.json does not exist yet. Last failure: task order-intake-cleanup/1 (from craftbook \"csv-order-cleanup\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.1,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 4.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 42.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 586772,
        "timing": {
          "startedAt": "2026-10-16T10:44:05.834Z",
          "finishedAt": "2026-10-16T10:53:52.606Z",
          "timeToFirstArtifactMs": 171209,
          "timeToLastArtifactWriteMs": 582128,
          "timeToFirstTokenMs": 12768,
          "firstTurnTtftMs": 10576,
          "timeToFirstToolCallMs": 17573
        },
        "toolCalls": 50,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 90,
        "prefillTokensPerSec": 1957.5,
        "decodeTokens": 48944,
        "peakRssMb": 16945,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.628Z",
        "trialId": "craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 1
      },
      "startedAt": "2026-10-16T10:54:01.911Z",
      "finishedAt": "2026-10-16T11:07:42.778Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: fraser/dc6e82fb exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `write_artifact` calls failed 3 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.1,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 90.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 820867,
        "timing": {
          "startedAt": "2026-10-16T10:54:01.911Z",
          "finishedAt": "2026-10-16T11:07:42.778Z",
          "timeToFirstArtifactMs": 316078,
          "timeToLastArtifactWriteMs": 794678,
          "timeToFirstTokenMs": 11895,
          "firstTurnTtftMs": 9821,
          "timeToFirstToolCallMs": 22968
        },
        "toolCalls": 65,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 102.32,
        "prefillTokensPerSec": 2174.69,
        "decodeTokens": 58821,
        "peakRssMb": 21309,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.629Z",
        "trialId": "craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 1
      },
      "startedAt": "2026-10-16T11:07:51.995Z",
      "finishedAt": "2026-10-16T11:16:11.792Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: reza/49f7c10b exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.9,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 1.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 16.7% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 499797,
        "timing": {
          "startedAt": "2026-10-16T11:07:51.995Z",
          "finishedAt": "2026-10-16T11:16:11.792Z",
          "timeToFirstArtifactMs": 204421,
          "timeToLastArtifactWriteMs": 204421,
          "timeToFirstTokenMs": 10887,
          "firstTurnTtftMs": 9292,
          "timeToFirstToolCallMs": 13810
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 101.64,
        "prefillTokensPerSec": 2650.18,
        "decodeTokens": 47322,
        "peakRssMb": 18488,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.629Z",
        "trialId": "craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 1
      },
      "startedAt": "2026-10-16T11:16:21.605Z",
      "finishedAt": "2026-10-16T11:42:58.390Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-params:5:targetnone:fr163wvm:rp0:rf0\" stuck for 15m — 3 re-writes (39 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.1,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 71.4% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1596785,
        "timing": {
          "startedAt": "2026-10-16T11:16:21.605Z",
          "finishedAt": "2026-10-16T11:42:58.390Z",
          "timeToFirstArtifactMs": 278182,
          "timeToLastArtifactWriteMs": 1498356,
          "timeToFirstTokenMs": 14462,
          "firstTurnTtftMs": 12373,
          "timeToFirstToolCallMs": 20894
        },
        "toolCalls": 110,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.84,
        "prefillTokensPerSec": 1878.87,
        "decodeTokens": 97638,
        "peakRssMb": 19359,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T12:25:33.630Z",
        "trialId": "craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 1
      },
      "startedAt": "2026-10-16T11:43:08.254Z",
      "finishedAt": "2026-10-16T12:22:24.496Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on craftbook: inventory report gate script; missing signals stayed unchanged: craftbook \"stock-health-report\" has no step gated by a scope:\"craftbook\" script — attach the custom check script to the build step's gate, task inventory-health-check/1 (from craftbook \"stock-health-report\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count., out/inventory-report.json does not exist yet. Last failure: craftbook \"stock-health-report\" has no step gated by a scope:\"craftbook\" script — attach the custom check script to the build step's gate",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 4,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 40% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2356242,
        "timing": {
          "startedAt": "2026-10-16T11:43:08.254Z",
          "finishedAt": "2026-10-16T12:22:24.496Z",
          "timeToFirstArtifactMs": 412500,
          "timeToLastArtifactWriteMs": 2350123,
          "timeToFirstTokenMs": 11156,
          "firstTurnTtftMs": 9574,
          "timeToFirstToolCallMs": 14776
        },
        "toolCalls": 69,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 79.87,
        "prefillTokensPerSec": 1699.47,
        "decodeTokens": 183555,
        "peakRssMb": 16911,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.164Z",
        "trialId": "craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc",
      "cell": {
        "id": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 1
      },
      "startedAt": "2026-10-16T12:22:38.615Z",
      "finishedAt": "2026-10-16T12:36:31.202Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: reza/8fc9fb01 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `craftbook_write` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 2.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 25% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 832587,
        "timing": {
          "startedAt": "2026-10-16T12:22:38.615Z",
          "finishedAt": "2026-10-16T12:36:31.202Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 18508,
          "firstTurnTtftMs": 16206,
          "timeToFirstToolCallMs": 28223
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 76.04,
        "prefillTokensPerSec": 1445.72,
        "decodeTokens": 59959,
        "peakRssMb": 16954,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.167Z",
        "trialId": "craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc",
        "cellId": "r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 2
      },
      "startedAt": "2026-10-16T12:36:44.609Z",
      "finishedAt": "2026-10-16T12:38:41.249Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 116640,
        "timing": {
          "startedAt": "2026-10-16T12:36:44.609Z",
          "finishedAt": "2026-10-16T12:38:41.249Z",
          "timeToFirstArtifactMs": 66212,
          "timeToLastArtifactWriteMs": 109495,
          "timeToFirstTokenMs": 8041,
          "firstTurnTtftMs": 5859,
          "timeToFirstToolCallMs": 16159
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 90.29,
        "prefillTokensPerSec": 2382.62,
        "decodeTokens": 7944,
        "peakRssMb": 17844,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.167Z",
        "trialId": "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 3107,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 4,
          "functionalCompleteness": 7,
          "codeQuality": 5,
          "polish": 4
        },
        "meanScore": 5,
        "justification": "Playable alternating turns, win/draw detection, reset and a persistent session scoreboard. The added win highlight is broken: checkWinner returns a symbol while highlightWin expects an index array. Plain fixed-size buttons and minimal layout limit presentation and polish.",
        "artifactSha256": "24f850b3423b5be26a55de9a4d05909a69a05756964960c5aa1d6ef38286ff78",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 2
      },
      "startedAt": "2026-10-16T12:39:23.096Z",
      "finishedAt": "2026-10-16T12:42:34.826Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 191730,
        "timing": {
          "startedAt": "2026-10-16T12:39:23.096Z",
          "finishedAt": "2026-10-16T12:42:34.826Z",
          "timeToFirstArtifactMs": 184559,
          "timeToLastArtifactWriteMs": 184578,
          "timeToFirstTokenMs": 11129,
          "firstTurnTtftMs": 9222,
          "timeToFirstToolCallMs": 50846
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 102.37,
        "prefillTokensPerSec": 2031.43,
        "decodeTokens": 15897,
        "peakRssMb": 24862,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.170Z",
        "trialId": "petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/workspace/pet-shop-website/index.html",
        "artifactBytes": 2707,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 6,
          "codeQuality": 6,
          "polish": 4
        },
        "meanScore": 5.5,
        "justification": "Coherent warm palette, logo reference and responsive six-product grid with names and prices. Product clicks work, but supplies incorrectly trigger ready-for-adoption alerts and clickable divs lack keyboard interaction. The requested shop is present with limited interaction polish.",
        "artifactSha256": "8d9c252eebf0e9df8c6ebe568b3ce3eb3fba162ae374d2cebb7d58bdc3cb1789",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 2
      },
      "startedAt": "2026-10-16T12:43:14.220Z",
      "finishedAt": "2026-10-16T12:47:04.792Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 230572,
        "timing": {
          "startedAt": "2026-10-16T12:43:14.220Z",
          "finishedAt": "2026-10-16T12:47:04.792Z",
          "timeToFirstArtifactMs": 220916,
          "timeToLastArtifactWriteMs": 220933,
          "timeToFirstTokenMs": 10206,
          "firstTurnTtftMs": 8495,
          "timeToFirstToolCallMs": 50480
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 99.19,
        "prefillTokensPerSec": 1976.43,
        "decodeTokens": 17208,
        "peakRssMb": 17766,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.171Z",
        "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 5327,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 4,
          "functionalCompleteness": 1,
          "codeQuality": 3,
          "polish": 2
        },
        "meanScore": 2.5,
        "justification": "A canvas, movement and score HUD exist, but createProjectile is never called: space prevents scrolling without firing. Movement is extremely slow; dead enemies remain in draw/collision loops and game-over does not stop play. No restart. Automated PASS does not establish playable combat.",
        "artifactSha256": "fa79afd4099b61f5dbe16275ad314b4b5c80c9948271a1328653dc3c55a9407e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 2
      },
      "startedAt": "2026-10-16T12:47:43.363Z",
      "finishedAt": "2026-10-16T12:50:49.480Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 186117,
        "timing": {
          "startedAt": "2026-10-16T12:47:43.363Z",
          "finishedAt": "2026-10-16T12:50:49.480Z",
          "timeToFirstArtifactMs": 24268,
          "timeToLastArtifactWriteMs": 183960,
          "timeToFirstTokenMs": 7480,
          "firstTurnTtftMs": 5684,
          "timeToFirstToolCallMs": 24273
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 100.52,
        "prefillTokensPerSec": 2090.09,
        "decodeTokens": 11923,
        "peakRssMb": 18113,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.172Z",
        "trialId": "schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 2
      },
      "startedAt": "2026-10-16T12:51:01.210Z",
      "finishedAt": "2026-10-16T12:57:18.547Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on src/machine.ts; missing signals stayed unchanged: vitest-green. Last failure: vitest run failed (exit 1): × submit moves draft to pending and returns the new state 3ms | × cancel is allowed from draft, pending, and paid 0ms | × delivered and cancelled are terminal 0ms | × a rejected send leaves the state unchanged 0ms | FAIL tests/machine.test.ts > order machine: transitions > submit moves draft to pending and returns the new state | AssertionError: expected 'draft' to be 'pending' // Object.is equality | FAIL tests/machine.test.ts > order machine: transitions > cancel is allowed from draft, pending, and paid | Error: invalid transition: event \"cancel\" is not allowed in state \"draft\"",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 377337,
        "timing": {
          "startedAt": "2026-10-16T12:51:01.210Z",
          "finishedAt": "2026-10-16T12:57:18.547Z",
          "timeToFirstArtifactMs": 25182,
          "timeToLastArtifactWriteMs": 376052,
          "timeToFirstTokenMs": 6934,
          "firstTurnTtftMs": 5161,
          "timeToFirstToolCallMs": 7333
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 96.96,
        "prefillTokensPerSec": 2103.24,
        "decodeTokens": 29627,
        "peakRssMb": 18562,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.173Z",
        "trialId": "failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 2
      },
      "startedAt": "2026-10-16T12:57:28.267Z",
      "finishedAt": "2026-10-16T13:03:45.766Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on lib/paginate.mjs; missing signals stayed unchanged: accept-passes. Last failure: node accept.mjs crashed (exit 1): file://<tmp>/gezel-eval-symptom-debug-3Kybb3/lib/paginate.mjs:13",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 66.7% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 377499,
        "timing": {
          "startedAt": "2026-10-16T12:57:28.267Z",
          "finishedAt": "2026-10-16T13:03:45.766Z",
          "timeToFirstArtifactMs": 114553,
          "timeToLastArtifactWriteMs": 373062,
          "timeToFirstTokenMs": 6960,
          "firstTurnTtftMs": 5228,
          "timeToFirstToolCallMs": 7490
        },
        "toolCalls": 29,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 101.89,
        "prefillTokensPerSec": 2326.46,
        "decodeTokens": 34219,
        "peakRssMb": 17859,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.173Z",
        "trialId": "symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:03:55.726Z",
      "finishedAt": "2026-10-16T13:19:52.443Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 956717,
        "timing": {
          "startedAt": "2026-10-16T13:03:55.726Z",
          "finishedAt": "2026-10-16T13:19:52.443Z",
          "timeToFirstArtifactMs": 156892,
          "timeToLastArtifactWriteMs": 951076,
          "timeToFirstTokenMs": 8078,
          "firstTurnTtftMs": 6482,
          "timeToFirstToolCallMs": 10968
        },
        "toolCalls": 42,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 105.43,
        "prefillTokensPerSec": 2304.74,
        "decodeTokens": 49428,
        "peakRssMb": 17761,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.173Z",
        "trialId": "data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:20:05.903Z",
      "finishedAt": "2026-10-16T13:27:18.954Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 433051,
        "timing": {
          "startedAt": "2026-10-16T13:20:05.903Z",
          "finishedAt": "2026-10-16T13:27:18.954Z",
          "timeToFirstArtifactMs": 304600,
          "timeToLastArtifactWriteMs": 428859,
          "timeToFirstTokenMs": 8852,
          "firstTurnTtftMs": 7072,
          "timeToFirstToolCallMs": 9881
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 87.2,
        "prefillTokensPerSec": 1643.91,
        "decodeTokens": 31881,
        "peakRssMb": 18607,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.174Z",
        "trialId": "incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:27:32.533Z",
      "finishedAt": "2026-10-16T13:28:19.878Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 47345,
        "timing": {
          "startedAt": "2026-10-16T13:27:32.533Z",
          "finishedAt": "2026-10-16T13:28:19.878Z",
          "timeToFirstArtifactMs": 39370,
          "timeToLastArtifactWriteMs": 43579,
          "timeToFirstTokenMs": 12429,
          "firstTurnTtftMs": 10623,
          "timeToFirstToolCallMs": 13274
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 89.1,
        "prefillTokensPerSec": 1829.57,
        "decodeTokens": 2654,
        "peakRssMb": 16782,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.174Z",
        "trialId": "ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:28:29.683Z",
      "finishedAt": "2026-10-16T13:29:22.726Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 53043,
        "timing": {
          "startedAt": "2026-10-16T13:28:29.683Z",
          "finishedAt": "2026-10-16T13:29:22.726Z",
          "timeToFirstArtifactMs": 51426,
          "timeToLastArtifactWriteMs": 51426,
          "timeToFirstTokenMs": 7509,
          "firstTurnTtftMs": 5744,
          "timeToFirstToolCallMs": 51432
        },
        "toolCalls": 1,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.94,
        "prefillTokensPerSec": 3471.32,
        "decodeTokens": 4030,
        "peakRssMb": 16785,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.175Z",
        "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:29:36.581Z",
      "finishedAt": "2026-10-16T13:38:35.244Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on synthesis.md; the same scenario failure stayed unchanged. Last failure: - only 3 substantively cited claim(s) — name the seeded source file as a backticked filename inside the sentence (for example `finance.csv`) for at least six separate claims; numbered references such as [3] and a Sources list do not count",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 538663,
        "timing": {
          "startedAt": "2026-10-16T13:29:36.581Z",
          "finishedAt": "2026-10-16T13:38:35.244Z",
          "timeToFirstArtifactMs": 184094,
          "timeToLastArtifactWriteMs": 534425,
          "timeToFirstTokenMs": 12685,
          "firstTurnTtftMs": 10890,
          "timeToFirstToolCallMs": 15036
        },
        "toolCalls": 45,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 85.71,
        "prefillTokensPerSec": 1657.27,
        "decodeTokens": 35743,
        "peakRssMb": 19874,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.176Z",
        "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v",
        "cellId": "r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:38:48.520Z",
      "finishedAt": "2026-10-16T13:40:32.080Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 103560,
        "timing": {
          "startedAt": "2026-10-16T13:38:48.520Z",
          "finishedAt": "2026-10-16T13:40:32.080Z",
          "timeToFirstArtifactMs": 97884,
          "timeToLastArtifactWriteMs": 97884,
          "timeToFirstTokenMs": 6754,
          "firstTurnTtftMs": 4813,
          "timeToFirstToolCallMs": 8746
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 100.42,
        "prefillTokensPerSec": 1803.84,
        "decodeTokens": 7905,
        "peakRssMb": 16841,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.177Z",
        "trialId": "constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:40:43.050Z",
      "finishedAt": "2026-10-16T13:53:23.059Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on records/attendees.csv; the same scenario failure stayed unchanged. Last failure: csv shape: CSV row 3 has 5 column(s), expected 6. Keep empty placeholders as adjacent commas so every row matches the header. Header order: id | full_name | email | company | ticket_type | registered_date. Row 3: A2 | otis@bramble.example | Bramble & Cole | standard | 2026-06-16",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 760009,
        "timing": {
          "startedAt": "2026-10-16T13:40:43.050Z",
          "finishedAt": "2026-10-16T13:53:23.059Z",
          "timeToFirstArtifactMs": 74398,
          "timeToLastArtifactWriteMs": 757901,
          "timeToFirstTokenMs": 7262,
          "firstTurnTtftMs": 5474,
          "timeToFirstToolCallMs": 8658
        },
        "toolCalls": 31,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 79.03,
        "prefillTokensPerSec": 1734.29,
        "decodeTokens": 53946,
        "peakRssMb": 18433,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.177Z",
        "trialId": "records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:53:41.310Z",
      "finishedAt": "2026-10-16T13:55:47.955Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 126645,
        "timing": {
          "startedAt": "2026-10-16T13:53:41.310Z",
          "finishedAt": "2026-10-16T13:55:47.955Z",
          "timeToFirstArtifactMs": 124055,
          "timeToLastArtifactWriteMs": 124055,
          "timeToFirstTokenMs": 12300,
          "firstTurnTtftMs": 9136,
          "timeToFirstToolCallMs": 49808
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 78.93,
        "prefillTokensPerSec": 1446.73,
        "decodeTokens": 8681,
        "peakRssMb": 16755,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.178Z",
        "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:56:06.599Z",
      "finishedAt": "2026-10-16T13:58:31.321Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 144722,
        "timing": {
          "startedAt": "2026-10-16T13:56:06.599Z",
          "finishedAt": "2026-10-16T13:58:31.321Z",
          "timeToFirstArtifactMs": 41715,
          "timeToLastArtifactWriteMs": 139157,
          "timeToFirstTokenMs": 14068,
          "firstTurnTtftMs": 10951,
          "timeToFirstToolCallMs": 18129
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 78.31,
        "prefillTokensPerSec": 1575.46,
        "decodeTokens": 2399,
        "peakRssMb": 16867,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.178Z",
        "trialId": "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1487,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 7,
          "grounding": 6,
          "actionability": 7,
          "brevity": 8
        },
        "meanScore": 7,
        "justification": "Clear decision, owners, dates and dependencies largely match the transcript. It confuses scheduling the annual-plan review by August 30 with holding it by that date, then calls that action unowned despite naming Jordan in the table. Otherwise concise and usable.",
        "artifactSha256": "c7b10682f4772c79620ef6c4d61c9dec23c50e8533bccf3b84730b700ec6be3c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 2
      },
      "startedAt": "2026-10-16T13:58:42.726Z",
      "finishedAt": "2026-10-16T14:03:24.942Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 282216,
        "timing": {
          "startedAt": "2026-10-16T13:58:42.726Z",
          "finishedAt": "2026-10-16T14:03:24.942Z",
          "timeToFirstArtifactMs": 68656,
          "timeToLastArtifactWriteMs": 180478,
          "timeToFirstTokenMs": 13294,
          "firstTurnTtftMs": 6639,
          "timeToFirstToolCallMs": 20790
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 92.52,
        "prefillTokensPerSec": 2108.92,
        "decodeTokens": 22311,
        "peakRssMb": 18520,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.179Z",
        "trialId": "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 2792,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 6
        },
        "meanScore": 6,
        "justification": "Purpose, inputs, formulas, months and acceptance checks form a usable scope; the example arithmetic reconciles. The net-profit label overstates a model that omits product cost and refunded revenue, while the treatment and level of churn remain ambiguous.",
        "artifactSha256": "c10ce9c0ff28a733c9ce36d4e5e684d647c70b75912841af9bb13eae77ce93fa",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-16T14:03:39.104Z",
      "finishedAt": "2026-10-16T14:05:47.108Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 128004,
        "timing": {
          "startedAt": "2026-10-16T14:03:39.104Z",
          "finishedAt": "2026-10-16T14:05:47.108Z",
          "timeToFirstArtifactMs": 45237,
          "timeToLastArtifactWriteMs": 126063,
          "timeToFirstTokenMs": 11676,
          "firstTurnTtftMs": 9819,
          "timeToFirstToolCallMs": 15121
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 103.92,
        "prefillTokensPerSec": 2219.1,
        "decodeTokens": 9498,
        "peakRssMb": 17457,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.181Z",
        "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 2
      },
      "startedAt": "2026-10-16T14:05:56.119Z",
      "finishedAt": "2026-10-16T14:07:24.373Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 88254,
        "timing": {
          "startedAt": "2026-10-16T14:05:56.119Z",
          "finishedAt": "2026-10-16T14:07:24.373Z",
          "timeToFirstArtifactMs": 82448,
          "timeToLastArtifactWriteMs": 82448,
          "timeToFirstTokenMs": 8770,
          "firstTurnTtftMs": 7126,
          "timeToFirstToolCallMs": 12674
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 118.28,
        "prefillTokensPerSec": 2299.35,
        "decodeTokens": 8538,
        "peakRssMb": 17200,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.181Z",
        "trialId": "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 772,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 7,
          "candor": 7,
          "usefulness": 3
        },
        "meanScore": 5.7,
        "justification": "Theme identifier, fonts, colors, page counts and unresolved Callout Emphasis match the supplied tool facts. It clearly acknowledges the unresolved style without inventing a transfer inventory, but supplies no concrete next action for resolving that limitation.",
        "artifactSha256": "e9619565e7c510bd357a40083f8293614263da2d79d1e40da6645907596d8f02",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 2
      },
      "startedAt": "2026-10-16T14:07:34.729Z",
      "finishedAt": "2026-10-16T14:28:57.679Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: nosipho/fde37bcb exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model kept trying to save factual text before researching, so the turn was stopped after 3 blocked writes. The next continuation will be directed to collect source evidence first.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.9,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 3.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 35.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1282950,
        "timing": {
          "startedAt": "2026-10-16T14:07:34.729Z",
          "finishedAt": "2026-10-16T14:28:57.679Z",
          "timeToFirstArtifactMs": 87489,
          "timeToLastArtifactWriteMs": 1192379,
          "timeToFirstTokenMs": 15227,
          "firstTurnTtftMs": 8598,
          "timeToFirstToolCallMs": 19649
        },
        "toolCalls": 72,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 111.11,
        "prefillTokensPerSec": 2886.03,
        "decodeTokens": 73439,
        "peakRssMb": 20646,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.182Z",
        "trialId": "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1867,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 5,
          "outline-parity": 0,
          "grounding": 7,
          "convertibility": 2
        },
        "meanScore": 3.5,
        "justification": "Metrics and next actions follow the supplied brief, but the final source lacks source-path citations. Its single H1 and subsequent H2 sections do not reproduce the six locked same-title H1 slides, breaking outline parity and slide-boundary readiness for conversion.",
        "artifactSha256": "1f13658af6484cc1095214650a29986e86d30c56a1fae87abff68d2da2e34740",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 2
      },
      "startedAt": "2026-10-16T14:29:10.348Z",
      "finishedAt": "2026-10-16T14:35:12.960Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: word-band, claim-calibration. Last failure: research brief is 691 words; required range is 700-1,500",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 84.6% of its measured signals."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 362612,
        "timing": {
          "startedAt": "2026-10-16T14:29:10.348Z",
          "finishedAt": "2026-10-16T14:35:12.960Z",
          "timeToFirstArtifactMs": 137837,
          "timeToLastArtifactWriteMs": 359532,
          "timeToFirstTokenMs": 7109,
          "firstTurnTtftMs": 5273,
          "timeToFirstToolCallMs": 21761
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 92.6,
        "prefillTokensPerSec": 1918.63,
        "decodeTokens": 28376,
        "peakRssMb": 18197,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.184Z",
        "trialId": "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5046,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 3,
          "citationQuality": 4,
          "calibration": 4
        },
        "meanScore": 3.8,
        "justification": "Basic chronology and engine context survive, but the brief invents publication details, a four-times length claim and detailed algorithm outcomes beyond the seeded cards. Source tags do not support those additions. First-programmer caveats coexist with an unqualified attribution, and synthesis is repetitive.",
        "artifactSha256": "6a9d4762b1c479bd0e360e27242925facedcc4d84e5372ee764a85e5100d0029",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 2
      },
      "startedAt": "2026-10-16T14:35:25.053Z",
      "finishedAt": "2026-10-16T15:15:58.694Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: meike/a0c22e77 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `invoke_craftbook` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2433641,
        "timing": {
          "startedAt": "2026-10-16T14:35:25.053Z",
          "finishedAt": "2026-10-16T15:15:58.694Z",
          "timeToFirstArtifactMs": 64847,
          "timeToLastArtifactWriteMs": 2410279,
          "timeToFirstTokenMs": 21102,
          "firstTurnTtftMs": 14047,
          "timeToFirstToolCallMs": 26809
        },
        "toolCalls": 142,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 85.29,
        "prefillTokensPerSec": 1897.68,
        "decodeTokens": 179602,
        "peakRssMb": 19776,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.185Z",
        "trialId": "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 2565,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 0,
          "prioritization": 1,
          "conflict-handling": 1,
          "realism": 1
        },
        "meanScore": 0.8,
        "justification": "Generic A-H labels replace the supplied event identities, and an extra wrap-up appears without evidence. Agenda times contradict claimed conflicts; proposed moves create further overlaps. Fixed daily focus blocks are not grounded in actual load. Calendar query/date provenance still needs a separate grader audit.",
        "artifactSha256": "566e69c9f2185a1c9db0c6389a3ef57813f67537f179f26b7e8c378298aeec1a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:16:11.705Z",
      "finishedAt": "2026-10-16T15:27:30.779Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 679074,
        "timing": {
          "startedAt": "2026-10-16T15:16:11.705Z",
          "finishedAt": "2026-10-16T15:27:30.779Z",
          "timeToFirstArtifactMs": 92977,
          "timeToLastArtifactWriteMs": 676354,
          "timeToFirstTokenMs": 16313,
          "firstTurnTtftMs": 9771,
          "timeToFirstToolCallMs": 22392
        },
        "toolCalls": 39,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 108.02,
        "prefillTokensPerSec": 2821.76,
        "decodeTokens": 66166,
        "peakRssMb": 19682,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.186Z",
        "trialId": "craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2143,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared axes safety/mapping/signals/containment describe mock-service deployment, while the requested artifact is a statistical A/B readout. No meaningful quality score is assigned until that evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:27:40.506Z",
      "finishedAt": "2026-10-16T15:32:54.920Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 314414,
        "timing": {
          "startedAt": "2026-10-16T15:27:40.506Z",
          "finishedAt": "2026-10-16T15:32:54.920Z",
          "timeToFirstArtifactMs": 44980,
          "timeToLastArtifactWriteMs": 263074,
          "timeToFirstTokenMs": 14024,
          "firstTurnTtftMs": 7482,
          "timeToFirstToolCallMs": 18381
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 109.65,
        "prefillTokensPerSec": 2561.58,
        "decodeTokens": 29204,
        "peakRssMb": 18717,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.186Z",
        "trialId": "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 5339,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 6,
          "grounding": 6,
          "structure": 7,
          "actionability": 6
        },
        "meanScore": 6.3,
        "justification": "All six seeded sources and numerical findings are retained in consistent summary/critique/relevance sections. Several authority and coverage claims exceed the cards, and the phrase Older source (2024 years old) is plainly wrong. Relevance is useful but evidence-strength comparisons are weak.",
        "artifactSha256": "27cb15bce654eb37af5713f8aaee12c1ec91786b6b01c3bb8f46c88d1eeda303",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:33:03.869Z",
      "finishedAt": "2026-10-16T15:36:07.110Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 183241,
        "timing": {
          "startedAt": "2026-10-16T15:33:03.869Z",
          "finishedAt": "2026-10-16T15:36:07.110Z",
          "timeToFirstArtifactMs": 47366,
          "timeToLastArtifactWriteMs": 128952,
          "timeToFirstTokenMs": 13859,
          "firstTurnTtftMs": 7268,
          "timeToFirstToolCallMs": 18699
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 122,
        "prefillTokensPerSec": 2819.75,
        "decodeTokens": 16280,
        "peakRssMb": 19280,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.187Z",
        "trialId": "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl",
        "cellId": "r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 2175,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 7,
          "structure": 7,
          "convertibility": 8
        },
        "meanScore": 7.3,
        "justification": "Pilot metrics, unresolved complaint and next actions are faithful in a clear summary, table and findings structure. Complete Markdown is suitable for DOCX conversion. Expanded recommendations go beyond the sparse brief, but no extra metric is invented.",
        "artifactSha256": "af7f06c3b55e06facc148b98dd34da17dcdd166a96f59d27c0b23212234e0010",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:36:16.086Z",
      "finishedAt": "2026-10-16T15:48:48.615Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 3 while the failing detail kept changing; latest checked file packages/core/src/spatial/Geohash.ts. Last failure: getGeohashPath behavioral check failed: getGeohashPath sampled only 16 points across ~2228 km — steps must derive from the true great-circle distance (~149 at 15 km/step)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 60% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 752529,
        "timing": {
          "startedAt": "2026-10-16T15:36:16.086Z",
          "finishedAt": "2026-10-16T15:48:48.615Z",
          "timeToFirstArtifactMs": 47635,
          "timeToLastArtifactWriteMs": 749360,
          "timeToFirstTokenMs": 7393,
          "firstTurnTtftMs": 5795,
          "timeToFirstToolCallMs": 8649
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 103.46,
        "prefillTokensPerSec": 2072.13,
        "decodeTokens": 61865,
        "peakRssMb": 23222,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.188Z",
        "trialId": "fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:49:00.679Z",
      "finishedAt": "2026-10-16T15:49:39.371Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 38692,
        "timing": {
          "startedAt": "2026-10-16T15:49:00.679Z",
          "finishedAt": "2026-10-16T15:49:39.371Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 11173,
          "firstTurnTtftMs": 9470,
          "timeToFirstToolCallMs": 20071
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 116.17,
        "prefillTokensPerSec": 2421.52,
        "decodeTokens": 2589,
        "peakRssMb": 16824,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.188Z",
        "trialId": "dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:49:48.885Z",
      "finishedAt": "2026-10-16T15:52:38.099Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 5 event(s) across 4 kind(s) routed and counted — team shape: 2 distinct non-meester session(s) (work was split)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 169214,
        "timing": {
          "startedAt": "2026-10-16T15:49:48.885Z",
          "finishedAt": "2026-10-16T15:52:38.099Z",
          "timeToFirstArtifactMs": 166719,
          "timeToLastArtifactWriteMs": 166769,
          "timeToFirstTokenMs": 8659,
          "firstTurnTtftMs": 7114,
          "timeToFirstToolCallMs": 25514
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 122.98,
        "prefillTokensPerSec": 2653.71,
        "decodeTokens": 18877,
        "peakRssMb": 17216,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.189Z",
        "trialId": "interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 2
      },
      "startedAt": "2026-10-16T15:52:50.084Z",
      "finishedAt": "2026-10-16T16:03:19.575Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: khadidiatou/c8c3e894 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_NO_PRODUCED_OUTPUT",
            "summary": "No authored artifact event or scored sniff output was observed."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 629491,
        "timing": {
          "startedAt": "2026-10-16T15:52:50.084Z",
          "finishedAt": "2026-10-16T16:03:19.575Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 14994,
          "firstTurnTtftMs": 8393,
          "timeToFirstToolCallMs": 17233
        },
        "toolCalls": 64,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 96.11,
        "prefillTokensPerSec": 2053.89,
        "decodeTokens": 51967,
        "peakRssMb": 18729,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.189Z",
        "trialId": "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "noArtifact",
        "comparisonEligible": false,
        "justification": "The exact artifact basename declared by this scenario was not found in the captured workspace or artifacts drawer."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 2
      },
      "startedAt": "2026-10-16T16:03:31.945Z",
      "finishedAt": "2026-10-16T16:13:15.504Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 583559,
        "timing": {
          "startedAt": "2026-10-16T16:03:31.945Z",
          "finishedAt": "2026-10-16T16:13:15.504Z",
          "timeToFirstArtifactMs": 113158,
          "timeToLastArtifactWriteMs": 340932,
          "timeToFirstTokenMs": 14893,
          "firstTurnTtftMs": 8268,
          "timeToFirstToolCallMs": 22684
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 99.83,
        "prefillTokensPerSec": 2329,
        "decodeTokens": 50353,
        "peakRssMb": 20358,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.189Z",
        "trialId": "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 7511,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 3
        },
        "meanScore": 4,
        "justification": "The staged review identifies real retryCount, status-code and missing-auth mismatches. However, it falsely says only one list endpoint exists and treats an optional property as required. Its proposed security YAML has the wrong shape, so corrected snippets are not dependable without another review.",
        "artifactSha256": "6ad22cd9df76de31f760a917d603c42bd7fdf2e85c3a14b7d94566f981c3becd",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 2
      },
      "startedAt": "2026-10-16T16:13:24.554Z",
      "finishedAt": "2026-10-16T16:19:21.115Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 356561,
        "timing": {
          "startedAt": "2026-10-16T16:13:24.554Z",
          "finishedAt": "2026-10-16T16:19:21.115Z",
          "timeToFirstArtifactMs": 102193,
          "timeToLastArtifactWriteMs": 352916,
          "timeToFirstTokenMs": 8054,
          "firstTurnTtftMs": 6340,
          "timeToFirstToolCallMs": 8892
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 98.23,
        "prefillTokensPerSec": 2222.92,
        "decodeTokens": 28829,
        "peakRssMb": 20070,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.190Z",
        "trialId": "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 5228,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 7,
          "systemic analysis": 6,
          "remediation": 5
        },
        "meanScore": 6,
        "justification": "Three real admin vulnerabilities are traced and grouped, with concrete parameterized SQL and argument-array remedies in JSON. However, allowing requireUser as an alternative to requireAdmin does not ensure admin authorization, and the report misstates safe routes as using requireAdmin.",
        "artifactSha256": "23d445b8ef20d66c6339a6a7e37307ef4776d0d9b07fb625ed0c6f44a7f5a31c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 2
      },
      "startedAt": "2026-10-16T16:19:30.725Z",
      "finishedAt": "2026-10-16T16:36:07.357Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 996632,
        "timing": {
          "startedAt": "2026-10-16T16:19:30.725Z",
          "finishedAt": "2026-10-16T16:36:07.357Z",
          "timeToFirstArtifactMs": 338133,
          "timeToLastArtifactWriteMs": 993361,
          "timeToFirstTokenMs": 9555,
          "firstTurnTtftMs": 7843,
          "timeToFirstToolCallMs": 12449
        },
        "toolCalls": 71,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 89.99,
        "prefillTokensPerSec": 1973.46,
        "decodeTokens": 79394,
        "peakRssMb": 20356,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.191Z",
        "trialId": "large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 2
      },
      "startedAt": "2026-10-16T16:36:17.072Z",
      "finishedAt": "2026-10-16T16:50:17.476Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on workspace output; missing signals stayed unchanged: unchanged fixture package.json differs from its seeded content — revert it byte-for-byte. This scenario grades a REVIEW: report the defect in your findings, do not fix it in the source.. Last failure: unchanged fixture package.json differs from its seeded content — revert it byte-for-byte. This scenario grades a REVIEW: report the defect in your findings, do not fix it in the source.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 96.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 840404,
        "timing": {
          "startedAt": "2026-10-16T16:36:17.072Z",
          "finishedAt": "2026-10-16T16:50:17.476Z",
          "timeToFirstArtifactMs": 95210,
          "timeToLastArtifactWriteMs": 837940,
          "timeToFirstTokenMs": 14188,
          "firstTurnTtftMs": 7617,
          "timeToFirstToolCallMs": 15452
        },
        "toolCalls": 107,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 123.69,
        "prefillTokensPerSec": 2884.8,
        "decodeTokens": 81741,
        "peakRssMb": 20002,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.191Z",
        "trialId": "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 1814,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 7,
          "Invariant discipline": 8,
          "Reconciled verification": 7
        },
        "meanScore": 7.3,
        "justification": "The seven active definition/re-export/import/call/log/test sites are reconciled with the intentional changelog exclusion. Three green tests and residual scanning are reported. One blanket claim of no old name anywhere conflicts with the explicitly retained changelog, but the mechanical rename is disciplined.",
        "artifactSha256": "772dfbe6cf6d26cb928a87c19cf2dc016ade2a39f7180e0b2ca06ec4cb0bbaa4",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 2
      },
      "startedAt": "2026-10-16T16:50:28.010Z",
      "finishedAt": "2026-10-16T18:06:49.973Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: hoang/507b4b7e exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `run_installed_script` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 96.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4581963,
        "timing": {
          "startedAt": "2026-10-16T16:50:28.010Z",
          "finishedAt": "2026-10-16T18:06:49.973Z",
          "timeToFirstArtifactMs": 315451,
          "timeToLastArtifactWriteMs": 4463821,
          "timeToFirstTokenMs": 14219,
          "firstTurnTtftMs": 7649,
          "timeToFirstToolCallMs": 22449
        },
        "toolCalls": 408,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 106.36,
        "prefillTokensPerSec": 2165.32,
        "decodeTokens": 314293,
        "peakRssMb": 27301,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.192Z",
        "trialId": "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/artifacts/crewlog/change-notes.md",
        "artifactBytes": 1434,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 3,
          "Reversible-stage execution": 3,
          "Genuine deduplication": 4
        },
        "meanScore": 3.3,
        "justification": "The selected change-notes artifact still says work is unstarted, while saved source actually shares normalizeName across both callers. Companion baselines contradict one another and do not establish green runs between stages. The original titleCaseName export is lost, so the extraction does not preserve the full public API.",
        "artifactSha256": "2b9d280924d8241217180263f43d87db427bd0f41b8bd794a9e1289deaa93c26",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 2
      },
      "startedAt": "2026-10-16T18:07:01.701Z",
      "finishedAt": "2026-10-16T18:24:11.244Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "craftbook-bug-fix-tdd stale no-write repair loop: tasks/1/fix-notes.md had unchanged content (758 bytes) for 155s — a re-emission of identical bytes counts as unchanged here after repair feedback and newer chat activity; still failing 23/28: tasks/1/fix-notes.md is missing required content: fix-notes sections in order — nothing in its 758 bytes matches /^##\\s+Problem[\\s\\S]*^##\\s+Change[\\s\\S]*^##\\s+Files touched[\\s\\S]*^##\\s+Regression coverage[\\s\\S]*^##\\s+Risk[\\s\\S]*^##\\s+How to verify/. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 82.1% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1029543,
        "timing": {
          "startedAt": "2026-10-16T18:07:01.701Z",
          "finishedAt": "2026-10-16T18:24:11.244Z",
          "timeToFirstArtifactMs": 69999,
          "timeToLastArtifactWriteMs": 983024,
          "timeToFirstTokenMs": 14809,
          "firstTurnTtftMs": 8238,
          "timeToFirstToolCallMs": 16604
        },
        "toolCalls": 93,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 102.89,
        "prefillTokensPerSec": 1937.46,
        "decodeTokens": 84697,
        "peakRssMb": 22490,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.193Z",
        "trialId": "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy",
        "cellId": "r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 766,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 6,
          "Fix minimality": 8,
          "Honest verification": 5
        },
        "meanScore": 6.3,
        "justification": "Saved source uses the minimal > to >= fix with a threshold regression test. Companion repro and validation notes quote red and green runs. They incorrectly treat a passing reproduction as sufficient to finish the whole workflow, without distinguishing uncompleted review/gates from verified code behavior.",
        "artifactSha256": "c9911dad28d4edcf5054a76df194d77c0f9ba0cb0a1fa009762385f787d1debc",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read with exact frozen axes; selected seeded inputs and saved companion outputs inspected for grounding. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 2
      },
      "startedAt": "2026-10-16T18:24:20.717Z",
      "finishedAt": "2026-10-16T18:24:57.200Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 36483,
        "timing": {
          "startedAt": "2026-10-16T18:24:20.717Z",
          "finishedAt": "2026-10-16T18:24:57.200Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 11360,
          "firstTurnTtftMs": 9790,
          "timeToFirstToolCallMs": 16124
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 112.04,
        "prefillTokensPerSec": 2353.26,
        "decodeTokens": 1877,
        "peakRssMb": 16824,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-16T19:09:41.194Z",
        "trialId": "craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": 0,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This non-HTML task declares no qualitative axes; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 2
      },
      "startedAt": "2026-10-16T18:25:07.077Z",
      "finishedAt": "2026-10-16T19:30:05.646Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: mila/c08880be exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3898569,
        "timing": {
          "startedAt": "2026-10-16T18:25:07.077Z",
          "finishedAt": "2026-10-16T19:30:05.646Z",
          "timeToFirstArtifactMs": 216834,
          "timeToLastArtifactWriteMs": 3892260,
          "timeToFirstTokenMs": 13433,
          "firstTurnTtftMs": 11800,
          "timeToFirstToolCallMs": 17656
        },
        "toolCalls": 245,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 90.71,
        "prefillTokensPerSec": 1783.5,
        "decodeTokens": 245730,
        "peakRssMb": 22825,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.303Z",
        "trialId": "craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 2
      },
      "startedAt": "2026-10-16T19:30:18.651Z",
      "finishedAt": "2026-10-16T19:37:25.149Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 426498,
        "timing": {
          "startedAt": "2026-10-16T19:30:18.651Z",
          "finishedAt": "2026-10-16T19:37:25.149Z",
          "timeToFirstArtifactMs": 47750,
          "timeToLastArtifactWriteMs": 419974,
          "timeToFirstTokenMs": 15074,
          "firstTurnTtftMs": 8084,
          "timeToFirstToolCallMs": 17850
        },
        "toolCalls": 45,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 97.72,
        "prefillTokensPerSec": 2201.07,
        "decodeTokens": 33709,
        "peakRssMb": 19772,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.304Z",
        "trialId": "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 958,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 1,
          "numbering continuity": 8,
          "professionalism": 4,
          "completeness": 3,
          "print-polish": 6
        },
        "meanScore": 4.4,
        "justification": "Invoice 042 rebills an already-invoiced $1,200 item, inflating the run from $5,415 to $6,615. Invoice 043 invents SKU identifiers and allocations. Sequence 042–044 and terms-based due dates are correct, and print styles exist, but the ledger remains unchanged and the summary contradicts its own Harbor amount.",
        "artifactSha256": "f8686e487f2f5cfdf4d7a505d2f46e700a5a94055bec3df8fcd0e64e70a202a9",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 2
      },
      "startedAt": "2026-10-16T19:37:44.030Z",
      "finishedAt": "2026-10-16T20:12:05.572Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (chatter-path): sniff \"craftbook-author-linear:6:targetnone:fr1w57pz4:rp0:rf0\" stuck for 10m — 35 turns without artifact writes (19 tool calls) after scoring artifact. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2061542,
        "timing": {
          "startedAt": "2026-10-16T19:37:44.030Z",
          "finishedAt": "2026-10-16T20:12:05.572Z",
          "timeToFirstArtifactMs": 164371,
          "timeToLastArtifactWriteMs": 1457399,
          "timeToFirstTokenMs": 15414,
          "firstTurnTtftMs": 12414,
          "timeToFirstToolCallMs": 20596
        },
        "toolCalls": 97,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 84.23,
        "prefillTokensPerSec": 1799.52,
        "decodeTokens": 105819,
        "peakRssMb": 19511,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.305Z",
        "trialId": "craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 2
      },
      "startedAt": "2026-10-16T20:12:18.993Z",
      "finishedAt": "2026-10-16T20:33:24.014Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: sarah/10e2a21b exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `write_artifact` calls failed 3 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 72.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1265021,
        "timing": {
          "startedAt": "2026-10-16T20:12:18.993Z",
          "finishedAt": "2026-10-16T20:33:24.014Z",
          "timeToFirstArtifactMs": 419558,
          "timeToLastArtifactWriteMs": 1121254,
          "timeToFirstTokenMs": 9232,
          "firstTurnTtftMs": 7138,
          "timeToFirstToolCallMs": 23057
        },
        "toolCalls": 75,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 87.68,
        "prefillTokensPerSec": 1681.26,
        "decodeTokens": 74725,
        "peakRssMb": 22248,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.309Z",
        "trialId": "craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 2
      },
      "startedAt": "2026-10-16T20:33:39.844Z",
      "finishedAt": "2026-10-16T20:52:42.914Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: reza/cdc6f04c exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.4,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 1.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 16.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1143070,
        "timing": {
          "startedAt": "2026-10-16T20:33:39.844Z",
          "finishedAt": "2026-10-16T20:52:42.914Z",
          "timeToFirstArtifactMs": 241429,
          "timeToLastArtifactWriteMs": 241429,
          "timeToFirstTokenMs": 14148,
          "firstTurnTtftMs": 12279,
          "timeToFirstToolCallMs": 18128
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 87.63,
        "prefillTokensPerSec": 1664.73,
        "decodeTokens": 96041,
        "peakRssMb": 17860,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.310Z",
        "trialId": "craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 2
      },
      "startedAt": "2026-10-16T20:52:54.796Z",
      "finishedAt": "2026-10-16T21:22:16.975Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "chat stalled for 303s — daemon reachable but issuing no model turns (soft digest stuck at bac56c38f710bcbb) (re-engage nudge was sent and ignored)",
        "failureMode": "chat-stalled",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 4.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 42.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1762179,
        "timing": {
          "startedAt": "2026-10-16T20:52:54.796Z",
          "finishedAt": "2026-10-16T21:22:16.975Z",
          "timeToFirstArtifactMs": 626722,
          "timeToLastArtifactWriteMs": 1457968,
          "timeToFirstTokenMs": 11316,
          "firstTurnTtftMs": 9222,
          "timeToFirstToolCallMs": 19339
        },
        "toolCalls": 77,
        "autoAnswers": 5,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 81.4,
        "prefillTokensPerSec": 1976.91,
        "decodeTokens": 86587,
        "peakRssMb": 17001,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.310Z",
        "trialId": "craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 2
      },
      "startedAt": "2026-10-16T21:22:28.031Z",
      "finishedAt": "2026-10-16T21:59:04.727Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: reza-2/899b0bd3 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `write_artifact` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 60% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2196696,
        "timing": {
          "startedAt": "2026-10-16T21:22:28.031Z",
          "finishedAt": "2026-10-16T21:59:04.727Z",
          "timeToFirstArtifactMs": 1103049,
          "timeToLastArtifactWriteMs": 1992063,
          "timeToFirstTokenMs": 10075,
          "firstTurnTtftMs": 8493,
          "timeToFirstToolCallMs": 14266
        },
        "toolCalls": 86,
        "autoAnswers": 6,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 73.44,
        "prefillTokensPerSec": 1555.45,
        "decodeTokens": 156423,
        "peakRssMb": 16930,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.310Z",
        "trialId": "craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq",
      "cell": {
        "id": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 2
      },
      "startedAt": "2026-10-16T21:59:16.126Z",
      "finishedAt": "2026-10-16T22:03:51.551Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: reza/0051de24 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `invoke_craftbook` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 2.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 25% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 275425,
        "timing": {
          "startedAt": "2026-10-16T21:59:16.126Z",
          "finishedAt": "2026-10-16T22:03:51.551Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 14352,
          "firstTurnTtftMs": 12261,
          "timeToFirstToolCallMs": 19180
        },
        "toolCalls": 30,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 96.6,
        "prefillTokensPerSec": 1857.7,
        "decodeTokens": 23069,
        "peakRssMb": 16948,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.310Z",
        "trialId": "craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq",
        "cellId": "r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:04:01.260Z",
      "finishedAt": "2026-10-16T22:06:40.134Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 158874,
        "timing": {
          "startedAt": "2026-10-16T22:04:01.260Z",
          "finishedAt": "2026-10-16T22:06:40.134Z",
          "timeToFirstArtifactMs": 107278,
          "timeToLastArtifactWriteMs": 151087,
          "timeToFirstTokenMs": 7873,
          "firstTurnTtftMs": 5795,
          "timeToFirstToolCallMs": 10034
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 99.42,
        "prefillTokensPerSec": 2610.24,
        "decodeTokens": 11278,
        "peakRssMb": 18095,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.310Z",
        "trialId": "tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/workspace/default/index.html",
        "artifactBytes": 3038,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 4,
          "functionalCompleteness": 6,
          "codeQuality": 6,
          "polish": 3
        },
        "meanScore": 4.8,
        "justification": "Two-player turns, win/draw detection and reset work in a legible nine-button grid. It never shows whose turn it is, and render removes the X/O classes without restoring them, so the declared symbol colors never appear. Fixed sizing and bare styling provide little polish.",
        "artifactSha256": "9279d6ce4f626f2e0cc294d0d0cbbde7db67e76d0d72c73ebc1f6d34caf7fb14",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:07:16.708Z",
      "finishedAt": "2026-10-16T22:09:05.858Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 109150,
        "timing": {
          "startedAt": "2026-10-16T22:07:16.708Z",
          "finishedAt": "2026-10-16T22:09:05.858Z",
          "timeToFirstArtifactMs": 104799,
          "timeToLastArtifactWriteMs": 104810,
          "timeToFirstTokenMs": 8920,
          "firstTurnTtftMs": 7365,
          "timeToFirstToolCallMs": 29729
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 118.81,
        "prefillTokensPerSec": 2710.29,
        "decodeTokens": 10127,
        "peakRssMb": 24870,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.310Z",
        "trialId": "petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/workspace/pet-shop-website/index.html",
        "artifactBytes": 3904,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 5,
          "functionalCompleteness": 5,
          "codeQuality": 6,
          "polish": 3
        },
        "meanScore": 4.8,
        "justification": "Readable cards and a working accumulating cart form a basic shop. The custom logo is repeated as every pet avatar with empty alt text rather than used as a recognizable brand mark. There is no cart removal or quantity handling; styling and content remain generic.",
        "artifactSha256": "587e1d565ebcf76bfe1233da1189ad62ce1f31491f3fd06c7a88a67a5f5c668c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:09:42.571Z",
      "finishedAt": "2026-10-16T22:18:20.284Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 517713,
        "timing": {
          "startedAt": "2026-10-16T22:09:42.571Z",
          "finishedAt": "2026-10-16T22:18:20.284Z",
          "timeToFirstArtifactMs": 131104,
          "timeToLastArtifactWriteMs": 469839,
          "timeToFirstTokenMs": 8604,
          "firstTurnTtftMs": 7076,
          "timeToFirstToolCallMs": 41885
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 113.17,
        "prefillTokensPerSec": 2120.31,
        "decodeTokens": 45968,
        "peakRssMb": 19220,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/workspace/default/index.html",
        "artifactBytes": 10258,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 5,
          "functionalCompleteness": 1,
          "codeQuality": 2,
          "polish": 2
        },
        "meanScore": 2.5,
        "justification": "The canvas has tank shapes, grid, health and score displays, but window.gameState is never assigned before input handlers use it. Bullets never advance or age, fire cooldowns never decrement, and tanks never aim. Combat is therefore broken despite an automated PASS.",
        "artifactSha256": "b1c51161e08d3e8f8aadb0e846a8f2b0aadf7f1b0f9cedd2ba29bf147b847637",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:18:55.077Z",
      "finishedAt": "2026-10-16T22:23:36.357Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 6 while the failing detail kept changing; latest checked file src/types.ts. Last failure: tsc-clean failed: <tmp>/gezel-eval-tsc-fIbBep/tsconfig.json(13,6): error TS1005: ':' expected.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 281280,
        "timing": {
          "startedAt": "2026-10-16T22:18:55.077Z",
          "finishedAt": "2026-10-16T22:23:36.357Z",
          "timeToFirstArtifactMs": 19309,
          "timeToLastArtifactWriteMs": 278892,
          "timeToFirstTokenMs": 6826,
          "firstTurnTtftMs": 5217,
          "timeToFirstToolCallMs": 19313
        },
        "toolCalls": 41,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 121.54,
        "prefillTokensPerSec": 2785.45,
        "decodeTokens": 22824,
        "peakRssMb": 19933,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:23:45.757Z",
      "finishedAt": "2026-10-16T22:33:27.034Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "chat stalled for 302s — daemon reachable but issuing no model turns (soft digest stuck at 66769642cab09753) (re-engage nudge was sent and ignored)",
        "failureMode": "chat-stalled",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 581277,
        "timing": {
          "startedAt": "2026-10-16T22:23:45.757Z",
          "finishedAt": "2026-10-16T22:33:27.034Z",
          "timeToFirstArtifactMs": 22706,
          "timeToLastArtifactWriteMs": 277524,
          "timeToFirstTokenMs": 6729,
          "firstTurnTtftMs": 5122,
          "timeToFirstToolCallMs": 7020
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 136.46,
        "prefillTokensPerSec": 4769.38,
        "decodeTokens": 2966,
        "peakRssMb": 17769,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:33:36.165Z",
      "finishedAt": "2026-10-16T22:34:43.450Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 67285,
        "timing": {
          "startedAt": "2026-10-16T22:33:36.165Z",
          "finishedAt": "2026-10-16T22:34:43.450Z",
          "timeToFirstArtifactMs": 64180,
          "timeToLastArtifactWriteMs": 64180,
          "timeToFirstTokenMs": 6767,
          "firstTurnTtftMs": 5196,
          "timeToFirstToolCallMs": 7512
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 164.19,
        "prefillTokensPerSec": 3964.17,
        "decodeTokens": 9257,
        "peakRssMb": 16655,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 3
      },
      "startedAt": "2026-10-16T22:34:52.726Z",
      "finishedAt": "2026-10-16T23:10:06.069Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 605s — model busy but not delivering (hard digest stuck at 7509f2696d2b9092)",
        "failureMode": "model-stuck",
        "failureClass": "infra",
        "includedInFixedModelAggregate": false,
        "failureClassRule": "scheduler-draft-deadlock"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.7,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 2.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 27.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2113343,
        "timing": {
          "startedAt": "2026-10-16T22:34:52.726Z",
          "finishedAt": "2026-10-16T23:10:06.069Z",
          "timeToFirstArtifactMs": 42228,
          "timeToLastArtifactWriteMs": 1506648,
          "timeToFirstTokenMs": 6985,
          "firstTurnTtftMs": 5264,
          "timeToFirstToolCallMs": 8052
        },
        "toolCalls": 75,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 108.8,
        "prefillTokensPerSec": 2175.48,
        "decodeTokens": 119397,
        "peakRssMb": 20765,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:03:31.777Z",
      "finishedAt": "2026-10-17T01:08:05.523Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 273746,
        "timing": {
          "startedAt": "2026-10-17T01:03:31.777Z",
          "finishedAt": "2026-10-17T01:08:05.523Z",
          "timeToFirstArtifactMs": 128683,
          "timeToLastArtifactWriteMs": 271234,
          "timeToFirstTokenMs": 8657,
          "firstTurnTtftMs": 6968,
          "timeToFirstToolCallMs": 10297
        },
        "toolCalls": 27,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 132.13,
        "prefillTokensPerSec": 2967.17,
        "decodeTokens": 29215,
        "peakRssMb": 18491,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:08:15.399Z",
      "finishedAt": "2026-10-17T01:08:52.855Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37456,
        "timing": {
          "startedAt": "2026-10-17T01:08:15.399Z",
          "finishedAt": "2026-10-17T01:08:52.855Z",
          "timeToFirstArtifactMs": 28460,
          "timeToLastArtifactWriteMs": 36058,
          "timeToFirstTokenMs": 11302,
          "firstTurnTtftMs": 9707,
          "timeToFirstToolCallMs": 11904
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 105.68,
        "prefillTokensPerSec": 2200.78,
        "decodeTokens": 2507,
        "peakRssMb": 16811,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:09:02.575Z",
      "finishedAt": "2026-10-17T01:10:00.209Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 57634,
        "timing": {
          "startedAt": "2026-10-17T01:09:02.575Z",
          "finishedAt": "2026-10-17T01:10:00.209Z",
          "timeToFirstArtifactMs": 51343,
          "timeToLastArtifactWriteMs": 51343,
          "timeToFirstTokenMs": 6889,
          "firstTurnTtftMs": 5303,
          "timeToFirstToolCallMs": 51348
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 111.28,
        "prefillTokensPerSec": 4585.03,
        "decodeTokens": 5247,
        "peakRssMb": 16785,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.311Z",
        "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:10:10.131Z",
      "finishedAt": "2026-10-17T01:16:45.163Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/777ca664 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 395032,
        "timing": {
          "startedAt": "2026-10-17T01:10:10.131Z",
          "finishedAt": "2026-10-17T01:16:45.163Z",
          "timeToFirstArtifactMs": 90646,
          "timeToLastArtifactWriteMs": 330445,
          "timeToFirstTokenMs": 8598,
          "firstTurnTtftMs": 7010,
          "timeToFirstToolCallMs": 11491
        },
        "toolCalls": 46,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 103.59,
        "prefillTokensPerSec": 2460.21,
        "decodeTokens": 34056,
        "peakRssMb": 18862,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.312Z",
        "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc",
        "cellId": "r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:16:54.947Z",
      "finishedAt": "2026-10-17T01:17:21.915Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 26968,
        "timing": {
          "startedAt": "2026-10-17T01:16:54.947Z",
          "finishedAt": "2026-10-17T01:17:21.915Z",
          "timeToFirstArtifactMs": 10321,
          "timeToLastArtifactWriteMs": 23409,
          "timeToFirstTokenMs": 6817,
          "firstTurnTtftMs": 5232,
          "timeToFirstToolCallMs": 10324
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 123.88,
        "prefillTokensPerSec": 2965.7,
        "decodeTokens": 1600,
        "peakRssMb": 16883,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.312Z",
        "trialId": "constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:17:30.947Z",
      "finishedAt": "2026-10-17T01:20:19.558Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 168611,
        "timing": {
          "startedAt": "2026-10-17T01:17:30.947Z",
          "finishedAt": "2026-10-17T01:20:19.558Z",
          "timeToFirstArtifactMs": 58887,
          "timeToLastArtifactWriteMs": 166833,
          "timeToFirstTokenMs": 7808,
          "firstTurnTtftMs": 6229,
          "timeToFirstToolCallMs": 9117
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 118.92,
        "prefillTokensPerSec": 2198.55,
        "decodeTokens": 18758,
        "peakRssMb": 16790,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.312Z",
        "trialId": "records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:20:29.069Z",
      "finishedAt": "2026-10-17T01:21:46.791Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 77722,
        "timing": {
          "startedAt": "2026-10-17T01:20:29.069Z",
          "finishedAt": "2026-10-17T01:21:46.791Z",
          "timeToFirstArtifactMs": 72508,
          "timeToLastArtifactWriteMs": 72508,
          "timeToFirstTokenMs": 7884,
          "firstTurnTtftMs": 6287,
          "timeToFirstToolCallMs": 30961
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 119.23,
        "prefillTokensPerSec": 2213.7,
        "decodeTokens": 7827,
        "peakRssMb": 16747,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.312Z",
        "trialId": "plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:21:56.513Z",
      "finishedAt": "2026-10-17T01:24:04.302Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 127789,
        "timing": {
          "startedAt": "2026-10-17T01:21:56.513Z",
          "finishedAt": "2026-10-17T01:24:04.302Z",
          "timeToFirstArtifactMs": 26812,
          "timeToLastArtifactWriteMs": 124426,
          "timeToFirstTokenMs": 8928,
          "firstTurnTtftMs": 7351,
          "timeToFirstToolCallMs": 11480
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 138.83,
        "prefillTokensPerSec": 3112.46,
        "decodeTokens": 2941,
        "peakRssMb": 16960,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.312Z",
        "trialId": "meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1876,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 7,
          "grounding": 6,
          "actionability": 7,
          "brevity": 7
        },
        "meanScore": 6.8,
        "justification": "Decisions, action owners, dates, dependency and unresolved rollback ownership are clear. It correctly says Jordan must schedule the review by August 30, but later asserts the review itself is scheduled for that date. The annual-plan question is repeated across several sections.",
        "artifactSha256": "99bc6d65cc44c1e7dd443ee07fec7b63b0321537d56cf403eb1577a2f5a2c193",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:24:13.269Z",
      "finishedAt": "2026-10-17T01:36:57.586Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: moritz/e11967a5 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model couldn't apply its file edit after 5 tries, so the turn was stopped. Try sending your message again, or switch to a more capable model in Settings.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 83.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 764317,
        "timing": {
          "startedAt": "2026-10-17T01:24:13.269Z",
          "finishedAt": "2026-10-17T01:36:57.586Z",
          "timeToFirstArtifactMs": 38299,
          "timeToLastArtifactWriteMs": 185938,
          "timeToFirstTokenMs": 12434,
          "firstTurnTtftMs": 5915,
          "timeToFirstToolCallMs": 21522
        },
        "toolCalls": 53,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 110.69,
        "prefillTokensPerSec": 2393.46,
        "decodeTokens": 71739,
        "peakRssMb": 18570,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.312Z",
        "trialId": "craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 2275,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 3
        },
        "meanScore": 4,
        "justification": "Purpose, drivers, formulas and acceptance criteria are well separated, but the sanity example is wrong: 22,500 minus 9,000, 2,000 and 3,500 is 8,000 operating profit and 6,000 after tax, not 8,500 and 6,375. Cumulative cash balance is also mislabeled as cash runway.",
        "artifactSha256": "6464ea1f68eb3d7fe6889a863683c5ba65fa0539623c95f05625d9ddfd4d20c4",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:37:06.979Z",
      "finishedAt": "2026-10-17T01:42:56.622Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/17d27893 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 84.6% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 349643,
        "timing": {
          "startedAt": "2026-10-17T01:37:06.979Z",
          "finishedAt": "2026-10-17T01:42:56.622Z",
          "timeToFirstArtifactMs": 184847,
          "timeToLastArtifactWriteMs": 304911,
          "timeToFirstTokenMs": 8514,
          "firstTurnTtftMs": 6926,
          "timeToFirstToolCallMs": 9518
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 104.14,
        "prefillTokensPerSec": 2656.62,
        "decodeTokens": 30591,
        "peakRssMb": 18361,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.313Z",
        "trialId": "conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:43:05.502Z",
      "finishedAt": "2026-10-17T01:45:13.352Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 127850,
        "timing": {
          "startedAt": "2026-10-17T01:43:05.502Z",
          "finishedAt": "2026-10-17T01:45:13.352Z",
          "timeToFirstArtifactMs": 99736,
          "timeToLastArtifactWriteMs": 124607,
          "timeToFirstTokenMs": 8732,
          "firstTurnTtftMs": 7095,
          "timeToFirstToolCallMs": 18720
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 117.15,
        "prefillTokensPerSec": 2358.65,
        "decodeTokens": 12075,
        "peakRssMb": 17208,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.313Z",
        "trialId": "docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 812,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 4,
          "candor": 6,
          "usefulness": 5
        },
        "meanScore": 5,
        "justification": "The returned theme values and page counts are accurate and Callout Emphasis is identified. However, the report invents that original formatting was retained because the palette lacks a match, and attributes the page reduction to tighter spacing without tool evidence. Keeping the final deck is actionable but leaves the conflict unresolved.",
        "artifactSha256": "78beb5268c60c53966094d4b0637a150781c4983fae9acca833eb64ef4cdf72a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:45:22.338Z",
      "finishedAt": "2026-10-17T01:57:22.012Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 719674,
        "timing": {
          "startedAt": "2026-10-17T01:45:22.338Z",
          "finishedAt": "2026-10-17T01:57:22.012Z",
          "timeToFirstArtifactMs": 44736,
          "timeToLastArtifactWriteMs": 607800,
          "timeToFirstTokenMs": 15302,
          "firstTurnTtftMs": 8742,
          "timeToFirstToolCallMs": 16623
        },
        "toolCalls": 73,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 107.13,
        "prefillTokensPerSec": 2567.67,
        "decodeTokens": 70060,
        "peakRssMb": 18955,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.313Z",
        "trialId": "craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 2012,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 7,
          "outline-parity": 8,
          "grounding": 8,
          "convertibility": 7
        },
        "meanScore": 7.5,
        "justification": "All eight locked titles appear as H1s in order, with source references and faithful pilot metrics/actions. The source is concise enough for conversion. Visual instructions and a placeholder for an unavailable 18-SKU list remain presentation notes rather than finished slide content.",
        "artifactSha256": "12489870b161f3b4c4ae4a97507f446117ddaef017bb3beb8c61b6d9e75d46f2",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 3
      },
      "startedAt": "2026-10-17T01:57:31.398Z",
      "finishedAt": "2026-10-17T02:05:22.781Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 471383,
        "timing": {
          "startedAt": "2026-10-17T01:57:31.398Z",
          "finishedAt": "2026-10-17T02:05:22.781Z",
          "timeToFirstArtifactMs": 131962,
          "timeToLastArtifactWriteMs": 466925,
          "timeToFirstTokenMs": 6649,
          "firstTurnTtftMs": 5005,
          "timeToFirstToolCallMs": 21825
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 100.68,
        "prefillTokensPerSec": 2019.83,
        "decodeTokens": 41584,
        "peakRssMb": 19289,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.314Z",
        "trialId": "wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5410,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 2,
          "citationQuality": 4,
          "calibration": 4
        },
        "meanScore": 3.5,
        "justification": "The chronology and debated-label caveat are present, but unsupported quotations, death dates, archive claims and algorithm details are attributed to sparse cards. It also assigns engine architecture to the Menabrea card. Source-by-source repetition outweighs cross-source synthesis, and several claims overstate the evidence.",
        "artifactSha256": "dd003979135699b56428ac7067acd99177d9538a56d582a01dcebab5165b7a13",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 3
      },
      "startedAt": "2026-10-17T02:05:32.036Z",
      "finishedAt": "2026-10-17T03:19:51.711Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: meike/a8b02147 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `invoke_craftbook` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 1.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4459675,
        "timing": {
          "startedAt": "2026-10-17T02:05:32.036Z",
          "finishedAt": "2026-10-17T03:19:51.711Z",
          "timeToFirstArtifactMs": 952813,
          "timeToLastArtifactWriteMs": 3822964,
          "timeToFirstTokenMs": 17177,
          "firstTurnTtftMs": 10651,
          "timeToFirstToolCallMs": 19152
        },
        "toolCalls": 334,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 98.11,
        "prefillTokensPerSec": 2196.59,
        "decodeTokens": 367871,
        "peakRssMb": 23519,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.314Z",
        "trialId": "craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 3067,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 0,
          "prioritization": 1,
          "conflict-handling": 0,
          "realism": 1
        },
        "meanScore": 0.5,
        "justification": "The supplied eight events are replaced by generic meetings. Tuesday is called gapless despite a 15-minute gap; the proposed fix moves a follow-up instead of either allegedly conflicting meeting. Monday has 90 minutes labeled two hours. Event coverage, workload and conflict decisions are not grounded in the calendar.",
        "artifactSha256": "5023c3f510682f1e06ca962a0e53397501fb89e1a907f32e06b77f937759c631",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:20:03.866Z",
      "finishedAt": "2026-10-17T03:26:38.890Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 395024,
        "timing": {
          "startedAt": "2026-10-17T03:20:03.866Z",
          "finishedAt": "2026-10-17T03:26:38.890Z",
          "timeToFirstArtifactMs": 74059,
          "timeToLastArtifactWriteMs": 393662,
          "timeToFirstTokenMs": 15987,
          "firstTurnTtftMs": 9407,
          "timeToFirstToolCallMs": 26748
        },
        "toolCalls": 38,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 111.14,
        "prefillTokensPerSec": 2625.23,
        "decodeTokens": 36780,
        "peakRssMb": 20050,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.314Z",
        "trialId": "craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2260,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:26:48.668Z",
      "finishedAt": "2026-10-17T03:30:44.089Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 235421,
        "timing": {
          "startedAt": "2026-10-17T03:26:48.668Z",
          "finishedAt": "2026-10-17T03:30:44.089Z",
          "timeToFirstArtifactMs": 42056,
          "timeToLastArtifactWriteMs": 184977,
          "timeToFirstTokenMs": 14273,
          "firstTurnTtftMs": 7706,
          "timeToFirstToolCallMs": 20709
        },
        "toolCalls": 30,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 109.3,
        "prefillTokensPerSec": 2620.52,
        "decodeTokens": 20573,
        "peakRssMb": 19105,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.314Z",
        "trialId": "craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 6216,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 4,
          "grounding": 3,
          "structure": 4,
          "actionability": 4
        },
        "meanScore": 3.8,
        "justification": "Most seeded counts survive, but Patel is misread: breaking even after 1,600 encounters becomes clinical effectiveness and outperformance. Peer-review status and further outcomes are invented. Sections are scannable, yet author/year headings omit full titles and journals, so these are incomplete APA citations.",
        "artifactSha256": "4e0a68684f8d04c3399a5d5fe886534a8cbd36fe9b8d68466bde8f52eaf62957",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:30:53.205Z",
      "finishedAt": "2026-10-17T03:35:13.183Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 259978,
        "timing": {
          "startedAt": "2026-10-17T03:30:53.205Z",
          "finishedAt": "2026-10-17T03:35:13.183Z",
          "timeToFirstArtifactMs": 43192,
          "timeToLastArtifactWriteMs": 175223,
          "timeToFirstTokenMs": 14177,
          "firstTurnTtftMs": 7613,
          "timeToFirstToolCallMs": 15876
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 112.29,
        "prefillTokensPerSec": 2515.84,
        "decodeTokens": 22916,
        "peakRssMb": 19759,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.315Z",
        "trialId": "craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5",
        "cellId": "r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1706,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 8,
          "structure": 7,
          "convertibility": 8
        },
        "meanScore": 7.7,
        "justification": "The source metrics, unresolved complaint and three actions are faithfully retained in a usable title, summary, table and findings structure. Markdown is complete for DOCX conversion. Small interpretations extend the actions, but no additional metric or unsupported numerical result appears.",
        "artifactSha256": "b1e8544f574fe8df38b941f71b096dc7580feb315cc67fff7312e12cdcdcdbdf",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:35:22.920Z",
      "finishedAt": "2026-10-17T03:48:56.341Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"fix-squisq-bugs:4:targetnone:frnone:rp0:rf0\" stuck for 8m — 5 re-writes (22 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.2,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 813421,
        "timing": {
          "startedAt": "2026-10-17T03:35:22.920Z",
          "finishedAt": "2026-10-17T03:48:56.341Z",
          "timeToFirstArtifactMs": 50746,
          "timeToLastArtifactWriteMs": 714989,
          "timeToFirstTokenMs": 6733,
          "firstTurnTtftMs": 5138,
          "timeToFirstToolCallMs": 7284
        },
        "toolCalls": 49,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 88.39,
        "prefillTokensPerSec": 1611.57,
        "decodeTokens": 47076,
        "peakRssMb": 22841,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.315Z",
        "trialId": "fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:49:05.821Z",
      "finishedAt": "2026-10-17T03:49:27.864Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 22043,
        "timing": {
          "startedAt": "2026-10-17T03:49:05.821Z",
          "finishedAt": "2026-10-17T03:49:27.864Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 8609,
          "firstTurnTtftMs": 7038,
          "timeToFirstToolCallMs": 12112
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 116.76,
        "prefillTokensPerSec": 3572.51,
        "decodeTokens": 1228,
        "peakRssMb": 16921,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.315Z",
        "trialId": "dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:49:36.759Z",
      "finishedAt": "2026-10-17T03:56:31.577Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 2 kind(s) routed and counted — team shape: 2 distinct non-meester session(s) (work was split)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 414818,
        "timing": {
          "startedAt": "2026-10-17T03:49:36.759Z",
          "finishedAt": "2026-10-17T03:56:31.577Z",
          "timeToFirstArtifactMs": 99123,
          "timeToLastArtifactWriteMs": 412523,
          "timeToFirstTokenMs": 7423,
          "firstTurnTtftMs": 5882,
          "timeToFirstToolCallMs": 30024
        },
        "toolCalls": 32,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 114.39,
        "prefillTokensPerSec": 2306.04,
        "decodeTokens": 32310,
        "peakRssMb": 18458,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.315Z",
        "trialId": "interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 3
      },
      "startedAt": "2026-10-17T03:56:40.634Z",
      "finishedAt": "2026-10-17T04:01:09.692Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 269058,
        "timing": {
          "startedAt": "2026-10-17T03:56:40.634Z",
          "finishedAt": "2026-10-17T04:01:09.692Z",
          "timeToFirstArtifactMs": 216460,
          "timeToLastArtifactWriteMs": 233730,
          "timeToFirstTokenMs": 12349,
          "firstTurnTtftMs": 5776,
          "timeToFirstToolCallMs": 13994
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 129.48,
        "prefillTokensPerSec": 2797.58,
        "decodeTokens": 28342,
        "peakRssMb": 18362,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.315Z",
        "trialId": "craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 2273,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 3
        },
        "meanScore": 4,
        "justification": "Concrete SQL-injection and pagination findings support request-changes. However, the review flags deliberately documented best-effort audit writes and recommends rethrowing, contrary to the seeded convention. Missing amount validation and swallowed config errors are downplayed as nits, weakening severity prioritization.",
        "artifactSha256": "a28692ac7ff6975019978380fdfc4e74f21826e20aa538abdb9310a29f48fef1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 3
      },
      "startedAt": "2026-10-17T04:01:19.170Z",
      "finishedAt": "2026-10-17T04:12:52.286Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 693116,
        "timing": {
          "startedAt": "2026-10-17T04:01:19.170Z",
          "finishedAt": "2026-10-17T04:12:52.286Z",
          "timeToFirstArtifactMs": 116649,
          "timeToLastArtifactWriteMs": 293782,
          "timeToFirstTokenMs": 12652,
          "firstTurnTtftMs": 6095,
          "timeToFirstToolCallMs": 19652
        },
        "toolCalls": 35,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 100.29,
        "prefillTokensPerSec": 2625.71,
        "decodeTokens": 58902,
        "peakRssMb": 18961,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.316Z",
        "trialId": "craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 6413,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 2
        },
        "meanScore": 3.5,
        "justification": "The review catches retryCount and create-status problems, but contradictory compatibility conclusions, misidentified list routes and malformed correction snippets make it unreliable. security: [] does not add authentication, and calling authentication additive/nonbreaking understates the client impact. Findings also exceed the stated counts.",
        "artifactSha256": "19c4598f3ee460b66cf7555742ffc56be529ad80f1dc1baf9a077ca3892f0526",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 3
      },
      "startedAt": "2026-10-17T04:13:01.459Z",
      "finishedAt": "2026-10-17T04:18:14.692Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 313233,
        "timing": {
          "startedAt": "2026-10-17T04:13:01.459Z",
          "finishedAt": "2026-10-17T04:18:14.692Z",
          "timeToFirstArtifactMs": 74073,
          "timeToLastArtifactWriteMs": 311387,
          "timeToFirstTokenMs": 7880,
          "firstTurnTtftMs": 6283,
          "timeToFirstToolCallMs": 8990
        },
        "toolCalls": 31,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 106.26,
        "prefillTokensPerSec": 2389.48,
        "decodeTokens": 26671,
        "peakRssMb": 19778,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.316Z",
        "trialId": "craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 6682,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 7,
          "systemic analysis": 7,
          "remediation": 5
        },
        "meanScore": 6.3,
        "justification": "All three admin vulnerabilities are grounded in concrete paths, safe exceptions are distinguished correctly, and common root cause/blast radius are clear. SQL and admin middleware fixes are specific. Shell remediation still suggests passing separate arguments to exec, which is the wrong API; database write impact is asserted without privilege evidence.",
        "artifactSha256": "99bacddbee12fd6b06d8e77b945616ad737b1ae95136370feefe648e9a921635",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 3
      },
      "startedAt": "2026-10-17T04:18:24.396Z",
      "finishedAt": "2026-10-17T05:03:10.778Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"large-pr-review:3:targetpr-review.md:frddkaei:rp0:rf0\" stuck for 8m — 4 re-writes (8 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_UNGRADED_AUTHORED_OUTPUT",
            "summary": "An authored output exists without a sniff verdict; quality is capped at the rubric midpoint."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2686382,
        "timing": {
          "startedAt": "2026-10-17T04:18:24.396Z",
          "finishedAt": "2026-10-17T05:03:10.778Z",
          "timeToFirstArtifactMs": 237124,
          "timeToLastArtifactWriteMs": 2484523,
          "timeToFirstTokenMs": 11903,
          "firstTurnTtftMs": 10192,
          "timeToFirstToolCallMs": 14732
        },
        "toolCalls": 76,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 95.3,
        "prefillTokensPerSec": 1936.19,
        "decodeTokens": 214519,
        "peakRssMb": 26844,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.316Z",
        "trialId": "large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 3
      },
      "startedAt": "2026-10-17T05:03:31.850Z",
      "finishedAt": "2026-10-17T05:12:48.125Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: aslan/e49c0181 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `run_installed_script` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 96.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 556275,
        "timing": {
          "startedAt": "2026-10-17T05:03:31.850Z",
          "finishedAt": "2026-10-17T05:12:48.125Z",
          "timeToFirstArtifactMs": 52015,
          "timeToLastArtifactWriteMs": 511751,
          "timeToFirstTokenMs": 12568,
          "firstTurnTtftMs": 6072,
          "timeToFirstToolCallMs": 13610
        },
        "toolCalls": 111,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 127.04,
        "prefillTokensPerSec": 2878.91,
        "decodeTokens": 59041,
        "peakRssMb": 19394,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.316Z",
        "trialId": "craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 1592,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 5,
          "Invariant discipline": 7,
          "Reconciled verification": 4
        },
        "meanScore": 5.3,
        "justification": "The mechanical rename and changelog exclusion are clear, and saved tests use loadUser. However, the five-site enumeration omits the test import and test call that were also changed, so found/changed/excluded counts do not reconcile with the actual project. Three green tests do not cure that audit gap.",
        "artifactSha256": "6065b87f147e476aae0eac6bcbb1ce75b18ab2f093a58694d75055e7642d2bec",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 3
      },
      "startedAt": "2026-10-17T05:12:57.475Z",
      "finishedAt": "2026-10-17T05:24:53.075Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 715600,
        "timing": {
          "startedAt": "2026-10-17T05:12:57.475Z",
          "finishedAt": "2026-10-17T05:24:53.075Z",
          "timeToFirstArtifactMs": 164347,
          "timeToLastArtifactWriteMs": 710184,
          "timeToFirstTokenMs": 12887,
          "firstTurnTtftMs": 6350,
          "timeToFirstToolCallMs": 19688
        },
        "toolCalls": 113,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 119,
        "prefillTokensPerSec": 2603.63,
        "decodeTokens": 73313,
        "peakRssMb": 20594,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.317Z",
        "trialId": "craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 1396,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 6,
          "Reversible-stage execution": 6,
          "Genuine deduplication": 7
        },
        "meanScore": 6.3,
        "justification": "Baseline, characterization coverage and green stage claims are documented. Saved code has one helper used by both callers and retains titleCaseName. Companion verification nevertheless misstates exports, and the notes omit evidence supporting each intermediate run. The final extraction is stronger than its audit trail.",
        "artifactSha256": "297d7d9c0159d943b25c3bd56475161ed545239599715c7e38cda3c2a9f9e578",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 3
      },
      "startedAt": "2026-10-17T05:25:02.510Z",
      "finishedAt": "2026-10-17T05:39:47.736Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 885226,
        "timing": {
          "startedAt": "2026-10-17T05:25:02.510Z",
          "finishedAt": "2026-10-17T05:39:47.736Z",
          "timeToFirstArtifactMs": 40435,
          "timeToLastArtifactWriteMs": 839244,
          "timeToFirstTokenMs": 14746,
          "firstTurnTtftMs": 8224,
          "timeToFirstToolCallMs": 16252
        },
        "toolCalls": 103,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 109.51,
        "prefillTokensPerSec": 2677.38,
        "decodeTokens": 85233,
        "peakRssMb": 21806,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.317Z",
        "trialId": "craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs",
        "cellId": "r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1813,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 4,
          "Fix minimality": 8,
          "Honest verification": 6
        },
        "meanScore": 6,
        "justification": "The saved fix is the correct single-character boundary change with a threshold regression test. Green output is quoted, but red evidence mostly asserts failure rather than quoting a failing run. Review repeats that claim, and the risk note wrongly says changing the threshold would recreate the same off-by-one defect.",
        "artifactSha256": "311bec41eee42467313d2a431d8b30a377601717df40351702303021789009f5",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 3
      },
      "startedAt": "2026-10-17T05:39:57.205Z",
      "finishedAt": "2026-10-17T05:40:27.842Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 30637,
        "timing": {
          "startedAt": "2026-10-17T05:39:57.205Z",
          "finishedAt": "2026-10-17T05:40:27.842Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 11246,
          "firstTurnTtftMs": 9676,
          "timeToFirstToolCallMs": 17613
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 119.68,
        "prefillTokensPerSec": 2383,
        "decodeTokens": 1322,
        "peakRssMb": 16816,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 3
      },
      "startedAt": "2026-10-17T05:40:37.224Z",
      "finishedAt": "2026-10-17T05:57:12.332Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: reza-2/f3a4a070 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 995108,
        "timing": {
          "startedAt": "2026-10-17T05:40:37.224Z",
          "finishedAt": "2026-10-17T05:57:12.332Z",
          "timeToFirstArtifactMs": 214159,
          "timeToLastArtifactWriteMs": 990505,
          "timeToFirstTokenMs": 13279,
          "firstTurnTtftMs": 11642,
          "timeToFirstToolCallMs": 15103
        },
        "toolCalls": 57,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.58,
        "prefillTokensPerSec": 1695.28,
        "decodeTokens": 87620,
        "peakRssMb": 17071,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 3
      },
      "startedAt": "2026-10-17T05:57:22.034Z",
      "finishedAt": "2026-10-17T06:03:25.833Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 363799,
        "timing": {
          "startedAt": "2026-10-17T05:57:22.034Z",
          "finishedAt": "2026-10-17T06:03:25.833Z",
          "timeToFirstArtifactMs": 40460,
          "timeToLastArtifactWriteMs": 356913,
          "timeToFirstTokenMs": 14235,
          "firstTurnTtftMs": 7719,
          "timeToFirstToolCallMs": 15745
        },
        "toolCalls": 51,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 118.12,
        "prefillTokensPerSec": 2613.74,
        "decodeTokens": 35253,
        "peakRssMb": 19793,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 532,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 3,
          "numbering continuity": 8,
          "professionalism": 4,
          "completeness": 5,
          "print-polish": 6
        },
        "meanScore": 5.2,
        "justification": "Totals reconcile to $5,415 and numbers 042–044 continue correctly, but Kestrel is split into unsupported $800/$900/$900 SKU amounts and Bluestem gets an invented postal address. Three invoices and print rules exist; the original ledger is still unupdated and sender/payment details are thin.",
        "artifactSha256": "69f1f6b3732f756be2d46b1ea0593629deb1cc407e8dad0a9dc4a638d7870d72",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 3
      },
      "startedAt": "2026-10-17T06:03:35.505Z",
      "finishedAt": "2026-10-17T06:17:27.548Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on craftbook: CSV Order Cleanup; missing signals stayed unchanged: task order-intake-cleanup/1 (from craftbook \"csv-order-cleanup\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count., notes/anomalies.md is 0 bytes, need ≥ 80, out/report.md is 0 bytes, need ≥ 80, out/orders.json does not exist yet. Last failure: task order-intake-cleanup/1 (from craftbook \"csv-order-cleanup\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.1,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 4.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 42.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 832043,
        "timing": {
          "startedAt": "2026-10-17T06:03:35.505Z",
          "finishedAt": "2026-10-17T06:17:27.548Z",
          "timeToFirstArtifactMs": 395770,
          "timeToLastArtifactWriteMs": 828000,
          "timeToFirstTokenMs": 14151,
          "firstTurnTtftMs": 12069,
          "timeToFirstToolCallMs": 19813
        },
        "toolCalls": 48,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 91.54,
        "prefillTokensPerSec": 1718.37,
        "decodeTokens": 71792,
        "peakRssMb": 16945,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 3
      },
      "startedAt": "2026-10-17T06:17:36.582Z",
      "finishedAt": "2026-10-17T06:41:01.436Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1404854,
        "timing": {
          "startedAt": "2026-10-17T06:17:36.582Z",
          "finishedAt": "2026-10-17T06:41:01.436Z",
          "timeToFirstArtifactMs": 272312,
          "timeToLastArtifactWriteMs": 1400413,
          "timeToFirstTokenMs": 11585,
          "firstTurnTtftMs": 9527,
          "timeToFirstToolCallMs": 19479
        },
        "toolCalls": 112,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 107.31,
        "prefillTokensPerSec": 2242.15,
        "decodeTokens": 105962,
        "peakRssMb": 23633,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 3
      },
      "startedAt": "2026-10-17T06:41:10.628Z",
      "finishedAt": "2026-10-17T07:01:24.082Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "chat stalled for 303s — daemon reachable but issuing no model turns (soft digest stuck at 6575f422500d6d5f) (re-engage nudge was sent and ignored)",
        "failureMode": "chat-stalled",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2.4,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 1.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 16.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1213454,
        "timing": {
          "startedAt": "2026-10-17T06:41:10.628Z",
          "finishedAt": "2026-10-17T07:01:24.082Z",
          "timeToFirstArtifactMs": 269474,
          "timeToLastArtifactWriteMs": 717780,
          "timeToFirstTokenMs": 10548,
          "firstTurnTtftMs": 8975,
          "timeToFirstToolCallMs": 14269
        },
        "toolCalls": 41,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 106.29,
        "prefillTokensPerSec": 2185.5,
        "decodeTokens": 36405,
        "peakRssMb": 17765,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 3
      },
      "startedAt": "2026-10-17T07:01:36.835Z",
      "finishedAt": "2026-10-17T07:16:58.580Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on craftbook: regional sales rollup; missing signals stayed unchanged: no reusable authored craftbook with at least 2 steps exists yet — author the regional sales rollup as a reusable craftbook template, not as ad-hoc chat work. Last failure: no reusable authored craftbook with at least 2 steps exists yet — author the regional sales rollup as a reusable craftbook template, not as ad-hoc chat work",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 921745,
        "timing": {
          "startedAt": "2026-10-17T07:01:36.835Z",
          "finishedAt": "2026-10-17T07:16:58.580Z",
          "timeToFirstArtifactMs": 177456,
          "timeToLastArtifactWriteMs": 919956,
          "timeToFirstTokenMs": 10545,
          "firstTurnTtftMs": 8466,
          "timeToFirstToolCallMs": 13872
        },
        "toolCalls": 88,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 106.31,
        "prefillTokensPerSec": 2233.13,
        "decodeTokens": 87048,
        "peakRssMb": 19779,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 3
      },
      "startedAt": "2026-10-17T07:17:08.699Z",
      "finishedAt": "2026-10-17T07:38:20.579Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: iris/4660bb05 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model couldn't apply its file edit after 5 tries, so the turn was stopped. Try sending your message again, or switch to a more capable model in Settings.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.2,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1271880,
        "timing": {
          "startedAt": "2026-10-17T07:17:08.699Z",
          "finishedAt": "2026-10-17T07:38:20.579Z",
          "timeToFirstArtifactMs": 153214,
          "timeToLastArtifactWriteMs": 1073829,
          "timeToFirstTokenMs": 13747,
          "firstTurnTtftMs": 12075,
          "timeToFirstToolCallMs": 16820
        },
        "toolCalls": 112,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 106.31,
        "prefillTokensPerSec": 2236.16,
        "decodeTokens": 100815,
        "peakRssMb": 22285,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.318Z",
        "trialId": "craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb",
      "cell": {
        "id": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "lfm2.5-24b-a2b-q4",
        "targetId": "lfm2.5-24b-a2b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 3
      },
      "startedAt": "2026-10-17T07:38:30.468Z",
      "finishedAt": "2026-10-17T08:22:48.766Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:7:targetnone:fr1tjwffw:rp0:rf0:m7\" stuck for 8m — 8 re-writes (35 tool calls) without sniff movement. Artifact produced but never reached success (retry-loop nudge was sent and ignored).",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 87.5% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2658298,
        "timing": {
          "startedAt": "2026-10-17T07:38:30.468Z",
          "finishedAt": "2026-10-17T08:22:48.766Z",
          "timeToFirstArtifactMs": 430154,
          "timeToLastArtifactWriteMs": 2627064,
          "timeToFirstTokenMs": 14080,
          "firstTurnTtftMs": 12002,
          "timeToFirstToolCallMs": 17801
        },
        "toolCalls": 168,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 131072,
          "totalTokens": 131072,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "lfm2.5-24b-a2b-q4"
        },
        "decodeTokensPerSec": 102.67,
        "prefillTokensPerSec": 2140.46,
        "decodeTokens": 149402,
        "peakRssMb": 21131,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb",
        "cellId": "r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "lfm2.5-24b-a2b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:23:24.461Z",
      "finishedAt": "2026-10-17T08:24:09.407Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 44946,
        "timing": {
          "startedAt": "2026-10-17T08:23:24.461Z",
          "finishedAt": "2026-10-17T08:24:09.407Z",
          "timeToFirstArtifactMs": 40132,
          "timeToLastArtifactWriteMs": 40136,
          "timeToFirstTokenMs": 12693,
          "firstTurnTtftMs": 10815,
          "timeToFirstToolCallMs": 21016
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 111.04,
        "prefillTokensPerSec": 1915.68,
        "decodeTokens": 2346,
        "peakRssMb": 30063,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.797Z",
        "trialId": "tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 4104,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 8,
          "polish": 6
        },
        "meanScore": 7,
        "justification": "The requested game is complete: clear turns, win/draw handling, disabled occupied/finished cells, winning-line highlight and clean reset. The cohesive dark palette and hover feedback improve the basic grid. Code is small and organized; layout remains fixed-width and empty cells lack descriptive accessible labels.",
        "artifactSha256": "ca63a91c890251b492abdf7ec900eb308a2abfd26ee47e9abe1e8044f66c588f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:24:44.683Z",
      "finishedAt": "2026-10-17T08:26:14.028Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 89345,
        "timing": {
          "startedAt": "2026-10-17T08:24:44.683Z",
          "finishedAt": "2026-10-17T08:26:14.028Z",
          "timeToFirstArtifactMs": 67117,
          "timeToLastArtifactWriteMs": 67122,
          "timeToFirstTokenMs": 10677,
          "firstTurnTtftMs": 9134,
          "timeToFirstToolCallMs": 14768
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 140.65,
        "prefillTokensPerSec": 2175.28,
        "decodeTokens": 8209,
        "peakRssMb": 39615,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.799Z",
        "trialId": "petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/workspace/pet-shop-website/index.html",
        "artifactBytes": 17798,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 8,
          "functionalCompleteness": 7,
          "codeQuality": 7,
          "polish": 6
        },
        "meanScore": 7,
        "justification": "A coherent responsive shop with custom-logo reference, services, product prices, mobile navigation and consistent styling. The newsletter only displays a success message despite promising an inbox offer, and several service links are placeholders. Fictional testimonials/contact details are presented as real, limiting finish quality.",
        "artifactSha256": "9f6c60c856f3894e6a92ae47044fd6b92026dcf48fff449457db32ea45616fc0",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:26:50.196Z",
      "finishedAt": "2026-10-17T08:27:40.189Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 49993,
        "timing": {
          "startedAt": "2026-10-17T08:26:50.196Z",
          "finishedAt": "2026-10-17T08:27:40.189Z",
          "timeToFirstArtifactMs": 44569,
          "timeToLastArtifactWriteMs": 44573,
          "timeToFirstTokenMs": 11685,
          "firstTurnTtftMs": 10137,
          "timeToFirstToolCallMs": 15797
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 141.58,
        "prefillTokensPerSec": 2170.89,
        "decodeTokens": 3765,
        "peakRssMb": 30170,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.800Z",
        "trialId": "tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 9387,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 2,
          "codeQuality": 4,
          "polish": 4
        },
        "meanScore": 4,
        "justification": "Movement, firing, enemy AI, grid, HUD and restart scaffolding exist. However, rectCollision requires width/height and bullet objects define neither; both bullet collision paths therefore fail. Enemies cannot be killed, score cannot rise and the player cannot die. Particles and game-over polish are consequently unreachable.",
        "artifactSha256": "59bdabe2afc070c0604c760feb7a027dfbb69b256b9cbda5ff2da464b4d2a5cf",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:28:16.382Z",
      "finishedAt": "2026-10-17T08:29:34.413Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 78031,
        "timing": {
          "startedAt": "2026-10-17T08:28:16.382Z",
          "finishedAt": "2026-10-17T08:29:34.413Z",
          "timeToFirstArtifactMs": 11602,
          "timeToLastArtifactWriteMs": 73031,
          "timeToFirstTokenMs": 8967,
          "firstTurnTtftMs": 7386,
          "timeToFirstToolCallMs": 11604
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 146.72,
        "prefillTokensPerSec": 2167.92,
        "decodeTokens": 3172,
        "peakRssMb": 35113,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:29:43.803Z",
      "finishedAt": "2026-10-17T08:30:06.773Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 22970,
        "timing": {
          "startedAt": "2026-10-17T08:29:43.803Z",
          "finishedAt": "2026-10-17T08:30:06.773Z",
          "timeToFirstArtifactMs": 17734,
          "timeToLastArtifactWriteMs": 17736,
          "timeToFirstTokenMs": 9856,
          "firstTurnTtftMs": 8227,
          "timeToFirstToolCallMs": 10443
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 151.2,
        "prefillTokensPerSec": 2088.6,
        "decodeTokens": 763,
        "peakRssMb": 29547,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:30:15.774Z",
      "finishedAt": "2026-10-17T08:31:02.896Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 47122,
        "timing": {
          "startedAt": "2026-10-17T08:30:15.774Z",
          "finishedAt": "2026-10-17T08:31:02.896Z",
          "timeToFirstArtifactMs": 42503,
          "timeToLastArtifactWriteMs": 42503,
          "timeToFirstTokenMs": 9974,
          "firstTurnTtftMs": 8372,
          "timeToFirstToolCallMs": 10639
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 123.83,
        "prefillTokensPerSec": 2063.58,
        "decodeTokens": 3933,
        "peakRssMb": 29270,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:31:11.855Z",
      "finishedAt": "2026-10-17T08:32:09.151Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 57296,
        "timing": {
          "startedAt": "2026-10-17T08:31:11.855Z",
          "finishedAt": "2026-10-17T08:32:09.151Z",
          "timeToFirstArtifactMs": 26369,
          "timeToLastArtifactWriteMs": 55064,
          "timeToFirstTokenMs": 10181,
          "firstTurnTtftMs": 8596,
          "timeToFirstToolCallMs": 11067
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 129.19,
        "prefillTokensPerSec": 1914.06,
        "decodeTokens": 3569,
        "peakRssMb": 31174,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:32:19.006Z",
      "finishedAt": "2026-10-17T08:36:17.107Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=miss; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.2,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 8.9,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 88.9% of its measured signals."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 238101,
        "timing": {
          "startedAt": "2026-10-17T08:32:19.006Z",
          "finishedAt": "2026-10-17T08:36:17.107Z",
          "timeToFirstArtifactMs": 211052,
          "timeToLastArtifactWriteMs": 236426,
          "timeToFirstTokenMs": 14486,
          "firstTurnTtftMs": 12897,
          "timeToFirstToolCallMs": 16266
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 78.14,
        "prefillTokensPerSec": 1340.85,
        "decodeTokens": 13666,
        "peakRssMb": 33667,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:36:26.863Z",
      "finishedAt": "2026-10-17T08:37:24.047Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 57184,
        "timing": {
          "startedAt": "2026-10-17T08:36:26.863Z",
          "finishedAt": "2026-10-17T08:37:24.047Z",
          "timeToFirstArtifactMs": 29049,
          "timeToLastArtifactWriteMs": 52754,
          "timeToFirstTokenMs": 16192,
          "firstTurnTtftMs": 14601,
          "timeToFirstToolCallMs": 16724
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 85.9,
        "prefillTokensPerSec": 1439.68,
        "decodeTokens": 2233,
        "peakRssMb": 30852,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:37:32.900Z",
      "finishedAt": "2026-10-17T08:38:10.000Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37100,
        "timing": {
          "startedAt": "2026-10-17T08:37:32.900Z",
          "finishedAt": "2026-10-17T08:38:10.000Z",
          "timeToFirstArtifactMs": 32199,
          "timeToLastArtifactWriteMs": 32199,
          "timeToFirstTokenMs": 10949,
          "firstTurnTtftMs": 9355,
          "timeToFirstToolCallMs": 11548
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 81.39,
        "prefillTokensPerSec": 1450.92,
        "decodeTokens": 1867,
        "peakRssMb": 29421,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:38:18.870Z",
      "finishedAt": "2026-10-17T08:40:16.335Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 117465,
        "timing": {
          "startedAt": "2026-10-17T08:38:18.870Z",
          "finishedAt": "2026-10-17T08:40:16.335Z",
          "timeToFirstArtifactMs": 69094,
          "timeToLastArtifactWriteMs": 113788,
          "timeToFirstTokenMs": 15056,
          "firstTurnTtftMs": 13452,
          "timeToFirstToolCallMs": 16034
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 101.98,
        "prefillTokensPerSec": 1628.78,
        "decodeTokens": 7156,
        "peakRssMb": 33476,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck",
        "cellId": "r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:40:25.235Z",
      "finishedAt": "2026-10-17T08:41:07.363Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 42128,
        "timing": {
          "startedAt": "2026-10-17T08:40:25.235Z",
          "finishedAt": "2026-10-17T08:41:07.363Z",
          "timeToFirstArtifactMs": 21843,
          "timeToLastArtifactWriteMs": 37952,
          "timeToFirstTokenMs": 9139,
          "firstTurnTtftMs": 7554,
          "timeToFirstToolCallMs": 11364
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 112.81,
        "prefillTokensPerSec": 1777.8,
        "decodeTokens": 1307,
        "peakRssMb": 31247,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:41:16.315Z",
      "finishedAt": "2026-10-17T08:42:43.659Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 87344,
        "timing": {
          "startedAt": "2026-10-17T08:41:16.315Z",
          "finishedAt": "2026-10-17T08:42:43.659Z",
          "timeToFirstArtifactMs": 36970,
          "timeToLastArtifactWriteMs": 84935,
          "timeToFirstTokenMs": 10602,
          "firstTurnTtftMs": 8998,
          "timeToFirstToolCallMs": 12206
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 115.76,
        "prefillTokensPerSec": 1532.5,
        "decodeTokens": 8209,
        "peakRssMb": 29551,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:42:53.506Z",
      "finishedAt": "2026-10-17T08:43:20.510Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 27004,
        "timing": {
          "startedAt": "2026-10-17T08:42:53.506Z",
          "finishedAt": "2026-10-17T08:43:20.510Z",
          "timeToFirstArtifactMs": 24855,
          "timeToLastArtifactWriteMs": 24855,
          "timeToFirstTokenMs": 10838,
          "firstTurnTtftMs": 9255,
          "timeToFirstToolCallMs": 11476
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 77.19,
        "prefillTokensPerSec": 1506.71,
        "decodeTokens": 1078,
        "peakRssMb": 29291,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.319Z",
        "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:43:29.467Z",
      "finishedAt": "2026-10-17T08:45:27.920Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 118453,
        "timing": {
          "startedAt": "2026-10-17T08:43:29.467Z",
          "finishedAt": "2026-10-17T08:45:27.920Z",
          "timeToFirstArtifactMs": 21774,
          "timeToLastArtifactWriteMs": 115341,
          "timeToFirstTokenMs": 12357,
          "firstTurnTtftMs": 10778,
          "timeToFirstToolCallMs": 13134
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 128.73,
        "prefillTokensPerSec": 1774.3,
        "decodeTokens": 1325,
        "peakRssMb": 30612,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.800Z",
        "trialId": "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1959,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 7,
          "grounding": 6,
          "actionability": 7,
          "brevity": 7
        },
        "meanScore": 6.8,
        "justification": "Decisions, rejected proposals, owners, dates and dependencies are clear, with explicit open questions. A4 wrongly requires Jordan to conduct the annual-plan review by August 30; the transcript only requires scheduling it by then. The open-question section correctly states the narrower commitment, creating an internal contradiction.",
        "artifactSha256": "3d92985cc362f7057afd9cc524c5310375ae6c657893befa08952a284bfff853",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:45:37.126Z",
      "finishedAt": "2026-10-17T08:51:02.513Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 325387,
        "timing": {
          "startedAt": "2026-10-17T08:45:37.126Z",
          "finishedAt": "2026-10-17T08:51:02.513Z",
          "timeToFirstArtifactMs": 37200,
          "timeToLastArtifactWriteMs": 133701,
          "timeToFirstTokenMs": 15613,
          "firstTurnTtftMs": 9113,
          "timeToFirstToolCallMs": 16793
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 93.7,
        "prefillTokensPerSec": 1927.19,
        "decodeTokens": 25903,
        "peakRssMb": 34097,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.801Z",
        "trialId": "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 6407,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 4,
          "usefulness": 2
        },
        "meanScore": 3,
        "justification": "Detailed drivers and formulas form a plausible scope, but acceptance numbers contradict the later correction. Even corrected arithmetic is wrong; positive monthly savings get N/A payback, and the claimed 539-order break-even conflicts with the formulas and the profitable 500-order baseline. This is not a reliable financial contract.",
        "artifactSha256": "8285d63f7462cf10b80167efd4033218508fa36c420f8e7c516c21c702f8335d",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:51:12.367Z",
      "finishedAt": "2026-10-17T08:54:55.454Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 223087,
        "timing": {
          "startedAt": "2026-10-17T08:51:12.367Z",
          "finishedAt": "2026-10-17T08:54:55.454Z",
          "timeToFirstArtifactMs": 34216,
          "timeToLastArtifactWriteMs": 218346,
          "timeToFirstTokenMs": 12857,
          "firstTurnTtftMs": 11273,
          "timeToFirstToolCallMs": 15578
        },
        "toolCalls": 46,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 93.59,
        "prefillTokensPerSec": 1378.77,
        "decodeTokens": 7231,
        "peakRssMb": 39996,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:55:05.210Z",
      "finishedAt": "2026-10-17T08:55:42.348Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37138,
        "timing": {
          "startedAt": "2026-10-17T08:55:05.210Z",
          "finishedAt": "2026-10-17T08:55:42.348Z",
          "timeToFirstArtifactMs": 31864,
          "timeToLastArtifactWriteMs": 34419,
          "timeToFirstTokenMs": 12131,
          "firstTurnTtftMs": 10495,
          "timeToFirstToolCallMs": 12626
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 103.68,
        "prefillTokensPerSec": 1339.58,
        "decodeTokens": 1643,
        "peakRssMb": 32041,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.801Z",
        "trialId": "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 955,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 2,
          "candor": 7,
          "usefulness": 6
        },
        "meanScore": 5,
        "justification": "Known theme tokens/page counts and the unresolved style are clear, with an actionable master-style recommendation. It invents a five-style transfer inventory, tighter margins as the page-count cause, retained source formatting and the absence of a branded accent. Those claims were not returned by the tools.",
        "artifactSha256": "3e98a259fca305007de5a32e5e40c947b40789b3f980d62af11ac0bace67ae23",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:55:51.202Z",
      "finishedAt": "2026-10-17T08:58:29.130Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 157928,
        "timing": {
          "startedAt": "2026-10-17T08:55:51.202Z",
          "finishedAt": "2026-10-17T08:58:29.130Z",
          "timeToFirstArtifactMs": 41152,
          "timeToLastArtifactWriteMs": 153204,
          "timeToFirstTokenMs": 19443,
          "firstTurnTtftMs": 12889,
          "timeToFirstToolCallMs": 19885
        },
        "toolCalls": 34,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 98.98,
        "prefillTokensPerSec": 1761.01,
        "decodeTokens": 8820,
        "peakRssMb": 35909,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.801Z",
        "trialId": "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1577,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 7,
          "outline-parity": 8,
          "grounding": 5,
          "convertibility": 8
        },
        "meanScore": 7,
        "justification": "All seven locked titles appear as H1s in order, with source references and preserved metrics/actions. Markdown is concise for conversion. Unsupported claims include why the SKUs were chosen, end-to-end cycle improvement and a process handoff rather than technology as the complaint cause. The narrative exceeds the source.",
        "artifactSha256": "b14e3a68b45b97ce59a3a36fac9c39e033a7befbbb88f17f8ee5d5cf2a57dbb1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 1
      },
      "startedAt": "2026-10-17T08:58:38.739Z",
      "finishedAt": "2026-10-17T09:01:01.510Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: claim-calibration. Last failure: every mention of the “first programmer” label must be hedged in the same sentence as debated shorthand",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 142771,
        "timing": {
          "startedAt": "2026-10-17T08:58:38.739Z",
          "finishedAt": "2026-10-17T09:01:01.510Z",
          "timeToFirstArtifactMs": 71460,
          "timeToLastArtifactWriteMs": 137622,
          "timeToFirstTokenMs": 9822,
          "firstTurnTtftMs": 8191,
          "timeToFirstToolCallMs": 29558
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 98.39,
        "prefillTokensPerSec": 1562.43,
        "decodeTokens": 8385,
        "peakRssMb": 33058,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.802Z",
        "trialId": "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5242,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 5,
          "grounding": 4,
          "citationQuality": 6,
          "calibration": 7
        },
        "meanScore": 5.5,
        "justification": "The three cards support the core chronology, design and carefully qualified first-programmer wording. However, the brief adds unsupported readership, textile-card history, loops/conditionals, century-scale architectural comparison and unpublished-work debate details. Nearby source IDs make supported facts easy to audit, but do not ground the added narrative.",
        "artifactSha256": "885f65f3054a0bea7d3bf25bd43d8a97f673f08b711abe1d40d7bedcef698362",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 1
      },
      "startedAt": "2026-10-17T09:01:11.293Z",
      "finishedAt": "2026-10-17T10:27:46.512Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-week-plan passed 12 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 5195219,
        "timing": {
          "startedAt": "2026-10-17T09:01:11.293Z",
          "finishedAt": "2026-10-17T10:27:46.512Z",
          "timeToFirstArtifactMs": 2356206,
          "timeToLastArtifactWriteMs": 5172490,
          "timeToFirstTokenMs": 22205,
          "firstTurnTtftMs": 15640,
          "timeToFirstToolCallMs": 24229
        },
        "toolCalls": 336,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 81.6,
        "prefillTokensPerSec": 1292.86,
        "decodeTokens": 331212,
        "peakRssMb": 45388,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.802Z",
        "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/workspace/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 5864,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 2,
          "prioritization": 2,
          "conflict-handling": 2,
          "realism": 1
        },
        "meanScore": 1.8,
        "justification": "Recognizable seeded meetings remain, but extra standups/vendor sync and altered event times distort the week. It invents Priya availability, misstates the Tuesday overlap, and creates a Wednesday conflict. Follow-up slots disagree across sections; advertised 15 focus hours do not sum to the table. Detail obscures an unreliable calendar.",
        "artifactSha256": "043e416dbb0429c011eeeb95ab0490409906faf022a4f707f7f869bc4de75453",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:27:55.721Z",
      "finishedAt": "2026-10-17T10:33:09.383Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 313662,
        "timing": {
          "startedAt": "2026-10-17T10:27:55.721Z",
          "finishedAt": "2026-10-17T10:33:09.383Z",
          "timeToFirstArtifactMs": 41942,
          "timeToLastArtifactWriteMs": 311202,
          "timeToFirstTokenMs": 21111,
          "firstTurnTtftMs": 14537,
          "timeToFirstToolCallMs": 21912
        },
        "toolCalls": 45,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 94.37,
        "prefillTokensPerSec": 1683.69,
        "decodeTokens": 13918,
        "peakRssMb": 36959,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2808,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:33:19.241Z",
      "finishedAt": "2026-10-17T10:35:13.210Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 113969,
        "timing": {
          "startedAt": "2026-10-17T10:33:19.241Z",
          "finishedAt": "2026-10-17T10:35:13.210Z",
          "timeToFirstArtifactMs": 24432,
          "timeToLastArtifactWriteMs": 72981,
          "timeToFirstTokenMs": 15686,
          "firstTurnTtftMs": 9206,
          "timeToFirstToolCallMs": 16462
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 126.59,
        "prefillTokensPerSec": 2340.6,
        "decodeTokens": 8266,
        "peakRssMb": 33475,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.802Z",
        "trialId": "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 7423,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 7,
          "grounding": 4,
          "structure": 7,
          "actionability": 6
        },
        "meanScore": 6,
        "justification": "All six records and numerical findings, including Patel break-even, are retained with clear annotation sections. Every citation adds a fake DOI placeholder. Study quality, comparison groups, cross-sectional design and data saturation are inferred as facts beyond the cards. Recommendations are sensible, but provenance and certainty need tightening.",
        "artifactSha256": "4372384ecc4d65231c864a029c0c8fc174bd10692f918f2dcf13943dc5a5efd0",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:35:22.983Z",
      "finishedAt": "2026-10-17T10:37:46.600Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 143617,
        "timing": {
          "startedAt": "2026-10-17T10:35:22.983Z",
          "finishedAt": "2026-10-17T10:37:46.600Z",
          "timeToFirstArtifactMs": 26636,
          "timeToLastArtifactWriteMs": 92717,
          "timeToFirstTokenMs": 16572,
          "firstTurnTtftMs": 9935,
          "timeToFirstToolCallMs": 18158
        },
        "toolCalls": 40,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 108.01,
        "prefillTokensPerSec": 1906.29,
        "decodeTokens": 8309,
        "peakRssMb": 36524,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.803Z",
        "trialId": "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65",
        "cellId": "r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 2703,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 6,
          "structure": 7,
          "convertibility": 8
        },
        "meanScore": 7,
        "justification": "Core metrics, complaint and three next actions are retained in a useful summary, table and sections. Markdown is ready for DOCX conversion. It speculates about triage, automation and validation causes, then asserts customers hear nothing until resolution; the source only reports status silence after photo submission.",
        "artifactSha256": "68858525e2b979b6984dfd843163f545588053fc08e4d98e3ef2ef908e08259b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:37:55.745Z",
      "finishedAt": "2026-10-17T10:38:53.286Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 57541,
        "timing": {
          "startedAt": "2026-10-17T10:37:55.745Z",
          "finishedAt": "2026-10-17T10:38:53.286Z",
          "timeToFirstArtifactMs": 32817,
          "timeToLastArtifactWriteMs": 54838,
          "timeToFirstTokenMs": 10496,
          "firstTurnTtftMs": 8883,
          "timeToFirstToolCallMs": 11471
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 111.14,
        "prefillTokensPerSec": 1377.6,
        "decodeTokens": 4130,
        "peakRssMb": 29536,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:39:02.829Z",
      "finishedAt": "2026-10-17T10:39:30.241Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 27412,
        "timing": {
          "startedAt": "2026-10-17T10:39:02.829Z",
          "finishedAt": "2026-10-17T10:39:30.241Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 16257,
          "firstTurnTtftMs": 14671,
          "timeToFirstToolCallMs": 18215
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 81.74,
        "prefillTokensPerSec": 1413.49,
        "decodeTokens": 570,
        "peakRssMb": 29566,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:39:39.758Z",
      "finishedAt": "2026-10-17T10:40:16.841Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 7 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37083,
        "timing": {
          "startedAt": "2026-10-17T10:39:39.758Z",
          "finishedAt": "2026-10-17T10:40:16.841Z",
          "timeToFirstArtifactMs": 32742,
          "timeToLastArtifactWriteMs": 35663,
          "timeToFirstTokenMs": 12707,
          "firstTurnTtftMs": 11158,
          "timeToFirstToolCallMs": 19699
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 109.99,
        "prefillTokensPerSec": 1752,
        "decodeTokens": 1645,
        "peakRssMb": 30217,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:40:25.725Z",
      "finishedAt": "2026-10-17T10:42:48.293Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 142568,
        "timing": {
          "startedAt": "2026-10-17T10:40:25.725Z",
          "finishedAt": "2026-10-17T10:42:48.293Z",
          "timeToFirstArtifactMs": 108580,
          "timeToLastArtifactWriteMs": 108580,
          "timeToFirstTokenMs": 17212,
          "firstTurnTtftMs": 10720,
          "timeToFirstToolCallMs": 17840
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 102.89,
        "prefillTokensPerSec": 1721.8,
        "decodeTokens": 8243,
        "peakRssMb": 33400,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.803Z",
        "trialId": "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3359,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 7
        },
        "meanScore": 7,
        "justification": "The review grounds both SQL-injection sites, resource cleanup failure and masked config errors in the diff, distinguishes intentional safe conventions, and gives concrete fixes. Request-changes is justified. Pagination is underweighted as minor, and targeted regression-test recommendations would make the review more actionable.",
        "artifactSha256": "ae2d869c88b16178ff5a5e59be128123739af16421cf74c40d64af03f04fa8f8",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:42:57.184Z",
      "finishedAt": "2026-10-17T10:49:26.109Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 388925,
        "timing": {
          "startedAt": "2026-10-17T10:42:57.184Z",
          "finishedAt": "2026-10-17T10:49:26.109Z",
          "timeToFirstArtifactMs": 124731,
          "timeToLastArtifactWriteMs": 331469,
          "timeToFirstTokenMs": 17897,
          "firstTurnTtftMs": 11374,
          "timeToFirstToolCallMs": 19202
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 87.01,
        "prefillTokensPerSec": 1406.5,
        "decodeTokens": 26443,
        "peakRssMb": 33023,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.803Z",
        "trialId": "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 14551,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 4
        },
        "meanScore": 5,
        "justification": "Real missing/extra routes, retryCount drift, auth and error-schema gaps are found across both route files. However, correction snippets repeatedly use invalid #/responses references, and the report falsely says verb paths break generators. Severity/count totals disagree, and compatibility checks are wrongly deferred until v2. Useful findings need cleanup before implementation.",
        "artifactSha256": "4d170a81fafa96e5f883ea6f2a1d58dda278ba9bf629490e3cae65257878dc59",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:49:35.719Z",
      "finishedAt": "2026-10-17T10:51:49.168Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 133449,
        "timing": {
          "startedAt": "2026-10-17T10:49:35.719Z",
          "finishedAt": "2026-10-17T10:51:49.168Z",
          "timeToFirstArtifactMs": 45798,
          "timeToLastArtifactWriteMs": 127999,
          "timeToFirstTokenMs": 13889,
          "firstTurnTtftMs": 12291,
          "timeToFirstToolCallMs": 15583
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 97.38,
        "prefillTokensPerSec": 1463.5,
        "decodeTokens": 7215,
        "peakRssMb": 32846,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.804Z",
        "trialId": "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 4470,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 6,
          "remediation": 8
        },
        "meanScore": 7.3,
        "justification": "Concrete source-to-sink paths and both missing admin guards are confirmed, while documented safe constructs are excluded. Report/JSON give execFile, placeholder SQL and requireAdmin fixes, with runtime limits stated. The systemic section treats missing authorization as the common cause of injection too, rather than separating independent unsafe-input patterns.",
        "artifactSha256": "7b67e34322eff33f5c7e4f46624467dff84afcb809fa49753a86b667197366d7",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 1
      },
      "startedAt": "2026-10-17T10:51:58.141Z",
      "finishedAt": "2026-10-17T11:02:34.176Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 636035,
        "timing": {
          "startedAt": "2026-10-17T10:51:58.141Z",
          "finishedAt": "2026-10-17T11:02:34.176Z",
          "timeToFirstArtifactMs": 59594,
          "timeToLastArtifactWriteMs": 631743,
          "timeToFirstTokenMs": 17187,
          "firstTurnTtftMs": 15489,
          "timeToFirstToolCallMs": 19986
        },
        "toolCalls": 29,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 56.15,
        "prefillTokensPerSec": 648.61,
        "decodeTokens": 15254,
        "peakRssMb": 40375,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:02:43.633Z",
      "finishedAt": "2026-10-17T11:07:24.789Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 281156,
        "timing": {
          "startedAt": "2026-10-17T11:02:43.633Z",
          "finishedAt": "2026-10-17T11:07:24.789Z",
          "timeToFirstArtifactMs": 44180,
          "timeToLastArtifactWriteMs": 276596,
          "timeToFirstTokenMs": 18011,
          "firstTurnTtftMs": 11477,
          "timeToFirstToolCallMs": 19414
        },
        "toolCalls": 103,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 101.71,
        "prefillTokensPerSec": 1559.26,
        "decodeTokens": 18794,
        "peakRssMb": 39427,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.804Z",
        "trialId": "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 2204,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 7,
          "Invariant discipline": 8,
          "Reconciled verification": 7
        },
        "meanScore": 7.3,
        "justification": "The seven real definition/re-export/import/call/log/test occurrences are enumerated, with explicit historical/specification exclusions and quoted green test output. Residual scanning and counts reconcile. One blanket statement about every original workspace occurrence conflicts with the immediately documented exclusions, but the mechanical audit is otherwise clear.",
        "artifactSha256": "7d043d7c0883788b54b812b27eac9c87f5cccac47a205d04109bb9aa220001ba",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:07:33.668Z",
      "finishedAt": "2026-10-17T11:24:12.988Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: qing/dc6f3f1e exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `write_artifact` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 30% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 999320,
        "timing": {
          "startedAt": "2026-10-17T11:07:33.668Z",
          "finishedAt": "2026-10-17T11:24:12.988Z",
          "timeToFirstArtifactMs": 30967,
          "timeToLastArtifactWriteMs": 984179,
          "timeToFirstTokenMs": 15955,
          "firstTurnTtftMs": 9408,
          "timeToFirstToolCallMs": 30969
        },
        "toolCalls": 81,
        "autoAnswers": 4,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 84.82,
        "prefillTokensPerSec": 1287.67,
        "decodeTokens": 69539,
        "peakRssMb": 39390,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "noArtifact",
        "comparisonEligible": false,
        "justification": "The exact declared artifact basename was absent from the captured workspace and artifacts drawer."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:24:22.789Z",
      "finishedAt": "2026-10-17T11:30:34.553Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 371764,
        "timing": {
          "startedAt": "2026-10-17T11:24:22.789Z",
          "finishedAt": "2026-10-17T11:30:34.553Z",
          "timeToFirstArtifactMs": 25174,
          "timeToLastArtifactWriteMs": 368411,
          "timeToFirstTokenMs": 18613,
          "firstTurnTtftMs": 12087,
          "timeToFirstToolCallMs": 20136
        },
        "toolCalls": 84,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 94.77,
        "prefillTokensPerSec": 1553.87,
        "decodeTokens": 24808,
        "peakRssMb": 38969,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.804Z",
        "trialId": "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1",
        "cellId": "r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 2027,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 8,
          "Fix minimality": 8,
          "Honest verification": 7
        },
        "meanScore": 7.7,
        "justification": "Companion notes quote a real failing threshold test, followed by six green tests including below/above-boundary cases; saved source has the minimal >= fix. Integration gaps are disclosed. The files-touched inventory omits the added tests, and verification relies on unit results rather than claiming end-to-end proof.",
        "artifactSha256": "ddc66c66b20fd9bcfa303829d43baad7fa0e71385499cb487a5f2a73e2637097",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:30:44.238Z",
      "finishedAt": "2026-10-17T11:31:16.030Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 31792,
        "timing": {
          "startedAt": "2026-10-17T11:30:44.238Z",
          "finishedAt": "2026-10-17T11:31:16.030Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 15977,
          "firstTurnTtftMs": 14405,
          "timeToFirstToolCallMs": 17969
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 90.15,
        "prefillTokensPerSec": 1481.31,
        "decodeTokens": 681,
        "peakRssMb": 29571,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:31:25.359Z",
      "finishedAt": "2026-10-17T11:40:09.469Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 524110,
        "timing": {
          "startedAt": "2026-10-17T11:31:25.359Z",
          "finishedAt": "2026-10-17T11:40:09.469Z",
          "timeToFirstArtifactMs": 227267,
          "timeToLastArtifactWriteMs": 493740,
          "timeToFirstTokenMs": 19634,
          "firstTurnTtftMs": 17995,
          "timeToFirstToolCallMs": 20832
        },
        "toolCalls": 59,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 74.13,
        "prefillTokensPerSec": 1380.92,
        "decodeTokens": 27496,
        "peakRssMb": 37998,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.320Z",
        "trialId": "craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:40:18.460Z",
      "finishedAt": "2026-10-17T11:44:30.572Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 252112,
        "timing": {
          "startedAt": "2026-10-17T11:40:18.460Z",
          "finishedAt": "2026-10-17T11:44:30.572Z",
          "timeToFirstArtifactMs": 28568,
          "timeToLastArtifactWriteMs": 245700,
          "timeToFirstTokenMs": 18074,
          "firstTurnTtftMs": 11575,
          "timeToFirstToolCallMs": 18632
        },
        "toolCalls": 49,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 111.71,
        "prefillTokensPerSec": 1591.32,
        "decodeTokens": 13717,
        "peakRssMb": 39402,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.805Z",
        "trialId": "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 829,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 6,
          "numbering continuity": 8,
          "professionalism": 3,
          "completeness": 5,
          "print-polish": 8
        },
        "meanScore": 6,
        "justification": "All three ledger amounts and total $5,415 reconcile, with numbers 042–044 and strong paper styling. Invoice 042 is issued November 17 but due November 16; postal/reference details and zero tax are invented. The ledger remains at 041 with billables uninvoiced. Attractive pages mask incorrect terms and unfinished bookkeeping.",
        "artifactSha256": "0fb91e036a501e807be5d37df9dc03869b434c3a923d5f910f9d185ed0754248",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 1
      },
      "startedAt": "2026-10-17T11:44:40.504Z",
      "finishedAt": "2026-10-17T12:05:04.944Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-linear:6:targetnone:fr1bjnkm5:rp0:rf0\" stuck for 8m — 4 re-writes (23 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1224440,
        "timing": {
          "startedAt": "2026-10-17T11:44:40.504Z",
          "finishedAt": "2026-10-17T12:05:04.944Z",
          "timeToFirstArtifactMs": 223452,
          "timeToLastArtifactWriteMs": 1123988,
          "timeToFirstTokenMs": 20449,
          "firstTurnTtftMs": 18367,
          "timeToFirstToolCallMs": 22255
        },
        "toolCalls": 72,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 66.82,
        "prefillTokensPerSec": 1224.01,
        "decodeTokens": 48222,
        "peakRssMb": 37852,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.321Z",
        "trialId": "craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 1
      },
      "startedAt": "2026-10-17T12:05:13.775Z",
      "finishedAt": "2026-10-17T12:08:27.210Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 193435,
        "timing": {
          "startedAt": "2026-10-17T12:05:13.775Z",
          "finishedAt": "2026-10-17T12:08:27.210Z",
          "timeToFirstArtifactMs": 54072,
          "timeToLastArtifactWriteMs": 174910,
          "timeToFirstTokenMs": 13004,
          "firstTurnTtftMs": 10956,
          "timeToFirstToolCallMs": 14712
        },
        "toolCalls": 36,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 112.3,
        "prefillTokensPerSec": 2118.6,
        "decodeTokens": 14859,
        "peakRssMb": 36928,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.321Z",
        "trialId": "craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 1
      },
      "startedAt": "2026-10-17T12:08:36.761Z",
      "finishedAt": "2026-10-17T12:42:07.158Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2010397,
        "timing": {
          "startedAt": "2026-10-17T12:08:36.761Z",
          "finishedAt": "2026-10-17T12:42:07.158Z",
          "timeToFirstArtifactMs": 230848,
          "timeToLastArtifactWriteMs": 2006047,
          "timeToFirstTokenMs": 20670,
          "firstTurnTtftMs": 18962,
          "timeToFirstToolCallMs": 24925
        },
        "toolCalls": 130,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 68.77,
        "prefillTokensPerSec": 1025.16,
        "decodeTokens": 41426,
        "peakRssMb": 40008,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.321Z",
        "trialId": "craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 1
      },
      "startedAt": "2026-10-17T12:42:16.317Z",
      "finishedAt": "2026-10-17T13:04:16.021Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1319704,
        "timing": {
          "startedAt": "2026-10-17T12:42:16.317Z",
          "finishedAt": "2026-10-17T13:04:16.021Z",
          "timeToFirstArtifactMs": 434612,
          "timeToLastArtifactWriteMs": 1314539,
          "timeToFirstTokenMs": 20347,
          "firstTurnTtftMs": 18251,
          "timeToFirstToolCallMs": 21531
        },
        "toolCalls": 77,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 71.24,
        "prefillTokensPerSec": 1072.89,
        "decodeTokens": 21775,
        "peakRssMb": 38431,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-17T13:19:08.321Z",
        "trialId": "craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 1
      },
      "startedAt": "2026-10-17T13:04:25.750Z",
      "finishedAt": "2026-10-17T13:59:26.520Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a craftbook with a substantive embedded gate script, the report parses, and the task completed",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3300770,
        "timing": {
          "startedAt": "2026-10-17T13:04:25.750Z",
          "finishedAt": "2026-10-17T13:59:26.520Z",
          "timeToFirstArtifactMs": 528933,
          "timeToLastArtifactWriteMs": 3050649,
          "timeToFirstTokenMs": 20114,
          "firstTurnTtftMs": 18540,
          "timeToFirstToolCallMs": 23036
        },
        "toolCalls": 79,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 65.82,
        "prefillTokensPerSec": 847.77,
        "decodeTokens": 56541,
        "peakRssMb": 45655,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.805Z",
        "trialId": "craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72",
      "cell": {
        "id": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 1
      },
      "startedAt": "2026-10-17T13:59:38.126Z",
      "finishedAt": "2026-10-17T14:26:40.538Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:6:targetnone:fr1tjwffw:rp0:rf0:m2\" stuck for 11m — 7 re-writes (30 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "infra",
        "includedInFixedModelAggregate": false,
        "failureClassRule": "gate-unevaluable"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1622412,
        "timing": {
          "startedAt": "2026-10-17T13:59:38.126Z",
          "finishedAt": "2026-10-17T14:26:40.538Z",
          "timeToFirstArtifactMs": 246921,
          "timeToLastArtifactWriteMs": 1619310,
          "timeToFirstTokenMs": 20469,
          "firstTurnTtftMs": 18377,
          "timeToFirstToolCallMs": 22387
        },
        "toolCalls": 73,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 65.17,
        "prefillTokensPerSec": 1020.25,
        "decodeTokens": 31964,
        "peakRssMb": 38895,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.805Z",
        "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72",
        "cellId": "r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:21:46.937Z",
      "finishedAt": "2026-10-18T00:22:22.042Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 35105,
        "timing": {
          "startedAt": "2026-10-18T00:21:46.937Z",
          "finishedAt": "2026-10-18T00:22:22.042Z",
          "timeToFirstArtifactMs": 27016,
          "timeToLastArtifactWriteMs": 27021,
          "timeToFirstTokenMs": 10890,
          "firstTurnTtftMs": 8914,
          "timeToFirstToolCallMs": 14545
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 159.33,
        "prefillTokensPerSec": 2735.96,
        "decodeTokens": 2033,
        "peakRssMb": 30277,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:23:44.805Z",
        "trialId": "tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 4134,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 8,
          "polish": 6
        },
        "meanScore": 7,
        "justification": "Turns, win/draw detection, guarded occupied cells and reset are complete. Symbol colors and a pulsing winning line add useful feedback to a cohesive dark layout. The fixed 312-pixel board is less adaptable on narrow screens, and continuous win animation has no reduced-motion handling; source organization is clear.",
        "artifactSha256": "c78209da7f1bc5582dfe81ac921ae4a9dc8bfaeeda82478fbe16ecbee70f91b2",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:22:57.301Z",
      "finishedAt": "2026-10-18T00:24:01.251Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 63950,
        "timing": {
          "startedAt": "2026-10-18T00:22:57.301Z",
          "finishedAt": "2026-10-18T00:24:01.251Z",
          "timeToFirstArtifactMs": 58046,
          "timeToLastArtifactWriteMs": 58053,
          "timeToFirstTokenMs": 10501,
          "firstTurnTtftMs": 8945,
          "timeToFirstToolCallMs": 13843
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 142.38,
        "prefillTokensPerSec": 2605.35,
        "decodeTokens": 6403,
        "peakRssMb": 38072,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.333Z",
        "trialId": "petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/workspace/pet-shop-website/index.html",
        "artifactBytes": 19256,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 7,
          "functionalCompleteness": 5,
          "codeQuality": 6,
          "polish": 5
        },
        "meanScore": 5.8,
        "justification": "Coherent responsive branding, product/service cards and a custom-logo reference meet the basic site brief. Added shopping buttons only link to #, and the contact form claims Message Sent without transmitting anything. The mobile menu works, but unlabelled form controls and invented reputation claims reduce professional finish.",
        "artifactSha256": "8177c42317dfe61b027a9404f0c08fced4a554185b428bde370a92db0434e6c0",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:24:37.807Z",
      "finishedAt": "2026-10-18T00:25:23.143Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 45336,
        "timing": {
          "startedAt": "2026-10-18T00:24:37.807Z",
          "finishedAt": "2026-10-18T00:25:23.143Z",
          "timeToFirstArtifactMs": 38020,
          "timeToLastArtifactWriteMs": 38025,
          "timeToFirstTokenMs": 10593,
          "firstTurnTtftMs": 9038,
          "timeToFirstToolCallMs": 14103
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 151.41,
        "prefillTokensPerSec": 2698.07,
        "decodeTokens": 3570,
        "peakRssMb": 30340,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 9856,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 7,
          "codeQuality": 6,
          "polish": 5
        },
        "meanScore": 6,
        "justification": "Movement, correctly sized projectile collisions, kills/score, contact defeat and restart form a playable loop with growing enemy pressure. Tank barrels are drawn a quarter-turn away from their firing direction; diagonal movement is faster and controls are not explained. Fixed canvas sizing and frame-based timing limit polish.",
        "artifactSha256": "14217f5cbae914fbdbd8be060c07aef9b2256bda6a43a2cbff181a572afec92f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:25:59.072Z",
      "finishedAt": "2026-10-18T00:27:18.153Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 79081,
        "timing": {
          "startedAt": "2026-10-18T00:25:59.072Z",
          "finishedAt": "2026-10-18T00:27:18.153Z",
          "timeToFirstArtifactMs": 11086,
          "timeToLastArtifactWriteMs": 74901,
          "timeToFirstTokenMs": 8780,
          "firstTurnTtftMs": 7188,
          "timeToFirstToolCallMs": 11088
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 141.15,
        "prefillTokensPerSec": 2233.6,
        "decodeTokens": 3881,
        "peakRssMb": 35306,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:27:28.003Z",
      "finishedAt": "2026-10-18T00:27:50.852Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 22849,
        "timing": {
          "startedAt": "2026-10-18T00:27:28.003Z",
          "finishedAt": "2026-10-18T00:27:50.852Z",
          "timeToFirstArtifactMs": 18890,
          "timeToLastArtifactWriteMs": 18892,
          "timeToFirstTokenMs": 10280,
          "firstTurnTtftMs": 8487,
          "timeToFirstToolCallMs": 11181
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 130.23,
        "prefillTokensPerSec": 1708.02,
        "decodeTokens": 645,
        "peakRssMb": 29400,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:28:00.161Z",
      "finishedAt": "2026-10-18T00:28:37.286Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37125,
        "timing": {
          "startedAt": "2026-10-18T00:28:00.161Z",
          "finishedAt": "2026-10-18T00:28:37.286Z",
          "timeToFirstArtifactMs": 35699,
          "timeToLastArtifactWriteMs": 35699,
          "timeToFirstTokenMs": 10451,
          "firstTurnTtftMs": 8857,
          "timeToFirstToolCallMs": 11260
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 98.59,
        "prefillTokensPerSec": 1645.74,
        "decodeTokens": 2387,
        "peakRssMb": 29156,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:28:47.610Z",
      "finishedAt": "2026-10-18T00:29:45.507Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 57897,
        "timing": {
          "startedAt": "2026-10-18T00:28:47.610Z",
          "finishedAt": "2026-10-18T00:29:45.507Z",
          "timeToFirstArtifactMs": 46149,
          "timeToLastArtifactWriteMs": 46149,
          "timeToFirstTokenMs": 10596,
          "firstTurnTtftMs": 9016,
          "timeToFirstToolCallMs": 11605
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 108.3,
        "prefillTokensPerSec": 1474.66,
        "decodeTokens": 4384,
        "peakRssMb": 29496,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:29:54.705Z",
      "finishedAt": "2026-10-18T00:33:32.652Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 217947,
        "timing": {
          "startedAt": "2026-10-18T00:29:54.705Z",
          "finishedAt": "2026-10-18T00:33:32.652Z",
          "timeToFirstArtifactMs": 189396,
          "timeToLastArtifactWriteMs": 213815,
          "timeToFirstTokenMs": 14977,
          "firstTurnTtftMs": 13395,
          "timeToFirstToolCallMs": 16540
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.57,
        "prefillTokensPerSec": 1361.48,
        "decodeTokens": 13025,
        "peakRssMb": 33334,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:33:42.535Z",
      "finishedAt": "2026-10-18T00:34:24.732Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 42197,
        "timing": {
          "startedAt": "2026-10-18T00:33:42.535Z",
          "finishedAt": "2026-10-18T00:34:24.732Z",
          "timeToFirstArtifactMs": 29603,
          "timeToLastArtifactWriteMs": 37993,
          "timeToFirstTokenMs": 15974,
          "firstTurnTtftMs": 14361,
          "timeToFirstToolCallMs": 16529
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 79.59,
        "prefillTokensPerSec": 1312.22,
        "decodeTokens": 1569,
        "peakRssMb": 29612,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.335Z",
        "trialId": "ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:34:34.540Z",
      "finishedAt": "2026-10-18T00:35:06.580Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 32040,
        "timing": {
          "startedAt": "2026-10-18T00:34:34.540Z",
          "finishedAt": "2026-10-18T00:35:06.580Z",
          "timeToFirstArtifactMs": 27346,
          "timeToLastArtifactWriteMs": 27346,
          "timeToFirstTokenMs": 10898,
          "firstTurnTtftMs": 9327,
          "timeToFirstToolCallMs": 11512
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.67,
        "prefillTokensPerSec": 1445.99,
        "decodeTokens": 1379,
        "peakRssMb": 29325,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.336Z",
        "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:35:15.476Z",
      "finishedAt": "2026-10-18T00:40:24.107Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/5f75f1cf exhausted its bounded automatic recovery allowance at the current checked progress; last error: [llama-cpp] prerequisite source-read repair exceeded its bounded read allowance before reading: memo-engineering.md, org.md, memo-oldplan.md",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 308631,
        "timing": {
          "startedAt": "2026-10-18T00:35:15.476Z",
          "finishedAt": "2026-10-18T00:40:24.107Z",
          "timeToFirstArtifactMs": 170215,
          "timeToLastArtifactWriteMs": 269893,
          "timeToFirstTokenMs": 15060,
          "firstTurnTtftMs": 13476,
          "timeToFirstToolCallMs": 16049
        },
        "toolCalls": 52,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 86.53,
        "prefillTokensPerSec": 1388.77,
        "decodeTokens": 16966,
        "peakRssMb": 38333,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.336Z",
        "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t",
        "cellId": "r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:40:33.678Z",
      "finishedAt": "2026-10-18T00:41:15.802Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 42124,
        "timing": {
          "startedAt": "2026-10-18T00:40:33.678Z",
          "finishedAt": "2026-10-18T00:41:15.802Z",
          "timeToFirstArtifactMs": 23890,
          "timeToLastArtifactWriteMs": 37546,
          "timeToFirstTokenMs": 9420,
          "firstTurnTtftMs": 7829,
          "timeToFirstToolCallMs": 12037
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 101.46,
        "prefillTokensPerSec": 1505.8,
        "decodeTokens": 1539,
        "peakRssMb": 31342,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.336Z",
        "trialId": "constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:41:24.677Z",
      "finishedAt": "2026-10-18T00:42:41.937Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 77260,
        "timing": {
          "startedAt": "2026-10-18T00:41:24.677Z",
          "finishedAt": "2026-10-18T00:42:41.937Z",
          "timeToFirstArtifactMs": 48033,
          "timeToLastArtifactWriteMs": 74096,
          "timeToFirstTokenMs": 10898,
          "firstTurnTtftMs": 9316,
          "timeToFirstToolCallMs": 12532
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 106.18,
        "prefillTokensPerSec": 1428.84,
        "decodeTokens": 6534,
        "peakRssMb": 29547,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.336Z",
        "trialId": "records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:42:50.813Z",
      "finishedAt": "2026-10-18T00:43:22.883Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 32070,
        "timing": {
          "startedAt": "2026-10-18T00:42:50.813Z",
          "finishedAt": "2026-10-18T00:43:22.883Z",
          "timeToFirstArtifactMs": 28991,
          "timeToLastArtifactWriteMs": 28991,
          "timeToFirstTokenMs": 10664,
          "firstTurnTtftMs": 9077,
          "timeToFirstToolCallMs": 11245
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.24,
        "prefillTokensPerSec": 1619.29,
        "decodeTokens": 1615,
        "peakRssMb": 29326,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:44:49.336Z",
        "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:43:31.750Z",
      "finishedAt": "2026-10-18T00:45:29.967Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 118217,
        "timing": {
          "startedAt": "2026-10-18T00:43:31.750Z",
          "finishedAt": "2026-10-18T00:45:29.967Z",
          "timeToFirstArtifactMs": 21848,
          "timeToLastArtifactWriteMs": 115419,
          "timeToFirstTokenMs": 12476,
          "firstTurnTtftMs": 10883,
          "timeToFirstToolCallMs": 13221
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 122.04,
        "prefillTokensPerSec": 1763.5,
        "decodeTokens": 1291,
        "peakRssMb": 30429,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:59:55.730Z",
        "trialId": "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1927,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 7,
          "grounding": 8,
          "actionability": 8,
          "brevity": 7
        },
        "meanScore": 7.5,
        "justification": "Decisions, superseded proposals, action owners/dates and the field-map dependency faithfully follow the transcript. It correctly keeps scheduling the annual-plan review distinct from deciding its scope, and leaves rollback ownership open. Follow-up is executable, though risks repeat material already stated in open questions.",
        "artifactSha256": "0a33b0b420dec5a34ab52f4618ef2c73d98946969db1b76a4ef8de01d7214002",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:45:39.382Z",
      "finishedAt": "2026-10-18T00:49:04.150Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 204768,
        "timing": {
          "startedAt": "2026-10-18T00:45:39.382Z",
          "finishedAt": "2026-10-18T00:49:04.150Z",
          "timeToFirstArtifactMs": 31566,
          "timeToLastArtifactWriteMs": 93558,
          "timeToFirstTokenMs": 15640,
          "firstTurnTtftMs": 9105,
          "timeToFirstToolCallMs": 17192
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 108.66,
        "prefillTokensPerSec": 2132.93,
        "decodeTokens": 17399,
        "peakRssMb": 31814,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:59:55.732Z",
        "trialId": "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 3510,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 4
        },
        "meanScore": 5,
        "justification": "Purpose, drivers, scenarios and acceptance checks provide a coherent scope, and the sanity arithmetic reconciles. However, the refund-amount input is unused, contradicting the promise that every input affects net savings. Response-time latency is mislabeled as labor hours saved, and the claimed records-derived $120 average is actually $119.",
        "artifactSha256": "7cf72ff3d5fd3cdda4d92b6059b9c3d9e8f01746f005016fae8c527ac6da513e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:49:13.181Z",
      "finishedAt": "2026-10-18T00:51:40.952Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 147771,
        "timing": {
          "startedAt": "2026-10-18T00:49:13.181Z",
          "finishedAt": "2026-10-18T00:51:40.952Z",
          "timeToFirstArtifactMs": 83776,
          "timeToLastArtifactWriteMs": 145546,
          "timeToFirstTokenMs": 14121,
          "firstTurnTtftMs": 12430,
          "timeToFirstToolCallMs": 15804
        },
        "toolCalls": 30,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 73.35,
        "prefillTokensPerSec": 1267.05,
        "decodeTokens": 6663,
        "peakRssMb": 34328,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:59:55.732Z",
        "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:51:49.735Z",
      "finishedAt": "2026-10-18T00:52:21.919Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 32184,
        "timing": {
          "startedAt": "2026-10-18T00:51:49.735Z",
          "finishedAt": "2026-10-18T00:52:21.919Z",
          "timeToFirstArtifactMs": 31025,
          "timeToLastArtifactWriteMs": 31025,
          "timeToFirstTokenMs": 10145,
          "firstTurnTtftMs": 8488,
          "timeToFirstToolCallMs": 10439
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 94.66,
        "prefillTokensPerSec": 1672.45,
        "decodeTokens": 1420,
        "peakRssMb": 31672,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:59:55.732Z",
        "trialId": "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1006,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 4,
          "candor": 7,
          "usefulness": 6
        },
        "meanScore": 5.7,
        "justification": "Exact theme tokens and page counts are retained, with a clear unresolved Callout Emphasis row and an actionable recommendation. It adds unsupported claims that no brand style matched, original formatting was retained and the apply tool confirmed skipping. Those details exceed the stated tool results.",
        "artifactSha256": "d5c5568821a0d3dba47da984ee3816b0d5b13aec0a03e3f8d2d63d3ede1314b5",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:52:31.712Z",
      "finishedAt": "2026-10-18T00:55:14.589Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 162877,
        "timing": {
          "startedAt": "2026-10-18T00:52:31.712Z",
          "finishedAt": "2026-10-18T00:55:14.589Z",
          "timeToFirstArtifactMs": 44536,
          "timeToLastArtifactWriteMs": 161330,
          "timeToFirstTokenMs": 20598,
          "firstTurnTtftMs": 14058,
          "timeToFirstToolCallMs": 21105
        },
        "toolCalls": 34,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 89.82,
        "prefillTokensPerSec": 1443.29,
        "decodeTokens": 8784,
        "peakRssMb": 35517,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T00:59:55.732Z",
        "trialId": "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1785,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 7,
          "outline-parity": 8,
          "grounding": 6,
          "convertibility": 8
        },
        "meanScore": 7.3,
        "justification": "All six locked outline titles appear once as H1s in order, with direct source references and faithful pilot metrics/actions. The source is concise for conversion. Claims about freeing agents from stalled-ticket follow-up, increased visibility and customer trust extend beyond the brief without being marked as interpretation.",
        "artifactSha256": "418375d0f669bc48dde47f9c4c550647cd3369af592e9bd07667cfb7a4b8bc53",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 2
      },
      "startedAt": "2026-10-18T00:55:24.297Z",
      "finishedAt": "2026-10-18T01:00:22.935Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: claim-calibration. Last failure: every mention of the “first programmer” label must be hedged in the same sentence as debated shorthand",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 298638,
        "timing": {
          "startedAt": "2026-10-18T00:55:24.297Z",
          "finishedAt": "2026-10-18T01:00:22.935Z",
          "timeToFirstArtifactMs": 111755,
          "timeToLastArtifactWriteMs": 293105,
          "timeToFirstTokenMs": 9996,
          "firstTurnTtftMs": 8353,
          "timeToFirstToolCallMs": 27020
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 92.81,
        "prefillTokensPerSec": 1354.03,
        "decodeTokens": 16167,
        "peakRssMb": 36665,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T01:14:35.311Z",
        "trialId": "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 7173,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 3,
          "citationQuality": 5,
          "calibration": 6
        },
        "meanScore": 4.5,
        "justification": "Core dates, design features and debated-label caveats follow the cards. Added claims about primary technical authority, readership, independence, branching and textile influence exceed them. Frequent source tags ground repeated facts but not those embellishments. The brief is padded and source-by-source, with awkward authorship wording.",
        "artifactSha256": "311e7c366037c1d44fb45edcb88fee915804494d036e8eeb64e6c586808165cf",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 2
      },
      "startedAt": "2026-10-18T01:00:32.551Z",
      "finishedAt": "2026-10-18T01:52:21.511Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "craftbook-week-plan stale no-write repair loop: tasks/eval/report.md had unchanged content (3716 bytes) for 61s — a re-emission of identical bytes counts as unchanged here after repair feedback and newer chat activity; still failing 8/12: tasks/eval/report.md is missing required content: the agenda-less vendor sync from the live calendar is in the plan — nothing in its 3716 bytes matches /Vendor sync/. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.4,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 66.7% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3108960,
        "timing": {
          "startedAt": "2026-10-18T01:00:32.551Z",
          "finishedAt": "2026-10-18T01:52:21.511Z",
          "timeToFirstArtifactMs": 1610835,
          "timeToLastArtifactWriteMs": 3095805,
          "timeToFirstTokenMs": 22807,
          "firstTurnTtftMs": 16258,
          "timeToFirstToolCallMs": 25290
        },
        "toolCalls": 254,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 77.42,
        "prefillTokensPerSec": 1261.13,
        "decodeTokens": 197021,
        "peakRssMb": 43005,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:00:21.437Z",
        "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/workspace/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 3812,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 0,
          "prioritization": 1,
          "conflict-handling": 1,
          "realism": 1
        },
        "meanScore": 0.8,
        "justification": "The retained calendar mock received zero requests, yet the report presents eight detailed meetings, times and a reschedule as established facts. Focus blocks and conflict handling therefore lack calendar grounding. It also calls nine listed focus blocks five, and describes both Tuesday meetings as 10am while later saying 10:30.",
        "artifactSha256": "221a71188955069080fff217ea26e1dbe2677bdec2f6a7a75b945836ff35f1ca",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 2
      },
      "startedAt": "2026-10-18T01:52:30.925Z",
      "finishedAt": "2026-10-18T02:04:31.144Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 720219,
        "timing": {
          "startedAt": "2026-10-18T01:52:30.925Z",
          "finishedAt": "2026-10-18T02:04:31.144Z",
          "timeToFirstArtifactMs": 38358,
          "timeToLastArtifactWriteMs": 715804,
          "timeToFirstTokenMs": 21376,
          "firstTurnTtftMs": 14857,
          "timeToFirstToolCallMs": 23378
        },
        "toolCalls": 86,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.98,
        "prefillTokensPerSec": 1475.4,
        "decodeTokens": 44776,
        "peakRssMb": 40369,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:14:29.450Z",
        "trialId": "craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 3486,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:04:41.113Z",
      "finishedAt": "2026-10-18T02:07:38.941Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 177828,
        "timing": {
          "startedAt": "2026-10-18T02:04:41.113Z",
          "finishedAt": "2026-10-18T02:07:38.941Z",
          "timeToFirstArtifactMs": 30146,
          "timeToLastArtifactWriteMs": 131855,
          "timeToFirstTokenMs": 18095,
          "firstTurnTtftMs": 11550,
          "timeToFirstToolCallMs": 19421
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 77.52,
        "prefillTokensPerSec": 1549,
        "decodeTokens": 8995,
        "peakRssMb": 33600,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:14:29.451Z",
        "trialId": "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 7490,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 7,
          "grounding": 5,
          "structure": 7,
          "actionability": 6
        },
        "meanScore": 6.3,
        "justification": "All six sources and seeded figures, including Patel break-even, are retained in consistent APA-style annotations. It adds ungrounded statistical-power judgments, comparison groups, geographic assumptions and author acknowledgments. Practical implications are useful, but evidence from the cards is not consistently separated from inference.",
        "artifactSha256": "84bbd1d3c88387f4066f5553b8c296e6262f31df5bfe56dd8cf8ef31ac0690ea",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:07:48.705Z",
      "finishedAt": "2026-10-18T02:09:52.098Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 123393,
        "timing": {
          "startedAt": "2026-10-18T02:07:48.705Z",
          "finishedAt": "2026-10-18T02:09:52.098Z",
          "timeToFirstArtifactMs": 29910,
          "timeToLastArtifactWriteMs": 65717,
          "timeToFirstTokenMs": 17885,
          "firstTurnTtftMs": 11327,
          "timeToFirstToolCallMs": 19981
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 90.69,
        "prefillTokensPerSec": 1549.04,
        "decodeTokens": 5477,
        "peakRssMb": 34297,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:14:29.452Z",
        "trialId": "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1",
        "cellId": "r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1462,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 5,
          "structure": 7,
          "convertibility": 8
        },
        "meanScore": 6.7,
        "justification": "Pilot metrics and required next actions survive in a clear summary, table and findings structure, with owner suggestions marked as hints. However, the source does not say customers hear nothing for days or that this causes refund requests. Those additions contradict the claim that every fact traces to the brief. Markdown is complete for DOCX conversion.",
        "artifactSha256": "12cbeeda9a197f44b4396bea3b4644657736e99e549fddbfbe7417198ca4ac16",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:10:01.055Z",
      "finishedAt": "2026-10-18T02:10:43.808Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 42753,
        "timing": {
          "startedAt": "2026-10-18T02:10:01.055Z",
          "finishedAt": "2026-10-18T02:10:43.808Z",
          "timeToFirstArtifactMs": 37070,
          "timeToLastArtifactWriteMs": 37096,
          "timeToFirstTokenMs": 10494,
          "firstTurnTtftMs": 8892,
          "timeToFirstToolCallMs": 11329
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 116.8,
        "prefillTokensPerSec": 1427.23,
        "decodeTokens": 2931,
        "peakRssMb": 29333,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:14:29.452Z",
        "trialId": "fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:10:53.030Z",
      "finishedAt": "2026-10-18T02:11:20.401Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 27371,
        "timing": {
          "startedAt": "2026-10-18T02:10:53.030Z",
          "finishedAt": "2026-10-18T02:11:20.401Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 15912,
          "firstTurnTtftMs": 14335,
          "timeToFirstToolCallMs": 17709
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 86.22,
        "prefillTokensPerSec": 1467.58,
        "decodeTokens": 639,
        "peakRssMb": 29581,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:14:29.452Z",
        "trialId": "dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:11:29.927Z",
      "finishedAt": "2026-10-18T02:23:43.302Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 4 while the failing detail kept changing; latest checked file src/consumer.ts. Last failure: node integration.mjs failed: sink.handle was called 1 time(s) for 5 event(s) — consume must route EVERY event to sink.handle(kind, payload)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.2,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 733375,
        "timing": {
          "startedAt": "2026-10-18T02:11:29.927Z",
          "finishedAt": "2026-10-18T02:23:43.302Z",
          "timeToFirstArtifactMs": 31364,
          "timeToLastArtifactWriteMs": 727953,
          "timeToFirstTokenMs": 12521,
          "firstTurnTtftMs": 10974,
          "timeToFirstToolCallMs": 21403
        },
        "toolCalls": 83,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 114.26,
        "prefillTokensPerSec": 1933.7,
        "decodeTokens": 26882,
        "peakRssMb": 38086,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:29:37.879Z",
        "trialId": "interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:23:52.644Z",
      "finishedAt": "2026-10-18T02:26:36.115Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 163471,
        "timing": {
          "startedAt": "2026-10-18T02:23:52.644Z",
          "finishedAt": "2026-10-18T02:26:36.115Z",
          "timeToFirstArtifactMs": 128824,
          "timeToLastArtifactWriteMs": 128824,
          "timeToFirstTokenMs": 16950,
          "firstTurnTtftMs": 10393,
          "timeToFirstToolCallMs": 17546
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 92.65,
        "prefillTokensPerSec": 1489.38,
        "decodeTokens": 9748,
        "peakRssMb": 32729,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:29:37.879Z",
        "trialId": "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3792,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 6
        },
        "meanScore": 5.5,
        "justification": "Concrete fixes cover SQL injection, masked config errors, leaked handles and pagination. However, the review raises best-effort audit writes despite an explicit house rule forbidding that finding, and wrongly calls failure silent although logger.warn reports it. Duplicate write-error commentary and uneven severity grading dilute an otherwise actionable review.",
        "artifactSha256": "18d19864ba1dbcbb46d599be5510e3b4c33f4279d5cd46a772f7b1125a9bfbd7",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:26:45.140Z",
      "finishedAt": "2026-10-18T02:34:14.844Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 449704,
        "timing": {
          "startedAt": "2026-10-18T02:26:45.140Z",
          "finishedAt": "2026-10-18T02:34:14.844Z",
          "timeToFirstArtifactMs": 124930,
          "timeToLastArtifactWriteMs": 391487,
          "timeToFirstTokenMs": 19050,
          "firstTurnTtftMs": 12536,
          "timeToFirstToolCallMs": 20642
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 78.28,
        "prefillTokensPerSec": 1267.64,
        "decodeTokens": 27864,
        "peakRssMb": 34937,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:44:45.107Z",
        "trialId": "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 11452,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 6
        },
        "meanScore": 6,
        "justification": "Real missing/extra routes, retryCount drift, status/error gaps and correct nullable-email handling are grounded, with mostly usable snippets. Finding totals and severity buckets disagree; a high issue is non-blocking. Claiming no versioning mechanism without /v1 overstates the evidence, and summary finding IDs do not match the headings.",
        "artifactSha256": "9cf435f75148f4986e8e34fc6d2a0e72448b335e114f663984308cc7663d9982",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:34:23.847Z",
      "finishedAt": "2026-10-18T02:36:32.631Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 128784,
        "timing": {
          "startedAt": "2026-10-18T02:34:23.847Z",
          "finishedAt": "2026-10-18T02:36:32.631Z",
          "timeToFirstArtifactMs": 46772,
          "timeToLastArtifactWriteMs": 124679,
          "timeToFirstTokenMs": 14198,
          "firstTurnTtftMs": 12598,
          "timeToFirstToolCallMs": 16491
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 92.12,
        "prefillTokensPerSec": 1437.9,
        "decodeTokens": 5747,
        "peakRssMb": 34791,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:44:45.108Z",
        "trialId": "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 5888,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 7,
          "systemic analysis": 6,
          "remediation": 6
        },
        "meanScore": 6.3,
        "justification": "All three admin vulnerabilities are traced, deliberate safe patterns are excluded, and SQL/admin fixes are concrete. The summary calls all three critical despite one high finding; database write access is asserted without privilege evidence. Shell fixes improve in JSON, but the report still allows validating input in a shell command contrary to its stated policy.",
        "artifactSha256": "bd53945322d984f84bc74ac68aa1eb2c9401129bfff6d113d1c31cca1596a790",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:36:42.310Z",
      "finishedAt": "2026-10-18T02:43:01.626Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 379316,
        "timing": {
          "startedAt": "2026-10-18T02:36:42.310Z",
          "finishedAt": "2026-10-18T02:43:01.626Z",
          "timeToFirstArtifactMs": 240310,
          "timeToLastArtifactWriteMs": 374433,
          "timeToFirstTokenMs": 17516,
          "firstTurnTtftMs": 15806,
          "timeToFirstToolCallMs": 19703
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 74.92,
        "prefillTokensPerSec": 934.45,
        "decodeTokens": 11800,
        "peakRssMb": 34952,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:44:45.109Z",
        "trialId": "large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:43:10.743Z",
      "finishedAt": "2026-10-18T02:53:21.304Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 610561,
        "timing": {
          "startedAt": "2026-10-18T02:43:10.743Z",
          "finishedAt": "2026-10-18T02:53:21.304Z",
          "timeToFirstArtifactMs": 56793,
          "timeToLastArtifactWriteMs": 605351,
          "timeToFirstTokenMs": 18355,
          "firstTurnTtftMs": 11823,
          "timeToFirstToolCallMs": 19910
        },
        "toolCalls": 79,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 84.02,
        "prefillTokensPerSec": 1455.04,
        "decodeTokens": 32795,
        "peakRssMb": 39918,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T02:59:42.249Z",
        "trialId": "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 1670,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 7,
          "Invariant discipline": 8,
          "Reconciled verification": 7
        },
        "meanScore": 7.3,
        "justification": "All seven definition/re-export/import/call/log/test sites are enumerated with defensible changelog/specification exclusions. Change notes and counts reconcile, three green tests are quoted, and the mechanical gate passed. The final residual scan is described over edited files rather than an explicit full-tree rerun, leaving a small audit-scope gap.",
        "artifactSha256": "e137c30b126d2c924b97e8e5801b06b9bfdbf60e5094a5f00d4fd0dfaf55ae7d",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 2
      },
      "startedAt": "2026-10-18T02:53:30.652Z",
      "finishedAt": "2026-10-18T03:23:30.114Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-refactor-module:25:targetnone:frvmrl5y:rp0:rf0:m11\" stuck for 8m — 6 re-writes (8 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 83.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1799462,
        "timing": {
          "startedAt": "2026-10-18T02:53:30.652Z",
          "finishedAt": "2026-10-18T03:23:30.114Z",
          "timeToFirstArtifactMs": 27190,
          "timeToLastArtifactWriteMs": 867845,
          "timeToFirstTokenMs": 14583,
          "firstTurnTtftMs": 8072,
          "timeToFirstToolCallMs": 27191
        },
        "toolCalls": 142,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 79.77,
        "prefillTokensPerSec": 1186.53,
        "decodeTokens": 116066,
        "peakRssMb": 41369,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T03:30:07.659Z",
        "trialId": "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 2059,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 1,
          "Reversible-stage execution": 1,
          "Genuine deduplication": 0
        },
        "meanScore": 0.7,
        "justification": "The notes audit a different API: two-name lowercasing and payroll reports replace the seeded title-casing/reportLine contract. Saved code omits reportLine/reportSummary, and the original verification fails. Sixteen green replacement tests do not establish a preserved baseline. Claims of no deviations and unchanged behavior are unsupported.",
        "artifactSha256": "0079e3a0c90d462e86196d91c11501a81265e9664cb48eae1096fd0175a85b5b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 2
      },
      "startedAt": "2026-10-18T03:23:39.841Z",
      "finishedAt": "2026-10-18T03:28:52.208Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 312367,
        "timing": {
          "startedAt": "2026-10-18T03:23:39.841Z",
          "finishedAt": "2026-10-18T03:28:52.208Z",
          "timeToFirstArtifactMs": 25600,
          "timeToLastArtifactWriteMs": 307344,
          "timeToFirstTokenMs": 18970,
          "firstTurnTtftMs": 12392,
          "timeToFirstToolCallMs": 20104
        },
        "toolCalls": 56,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 94.33,
        "prefillTokensPerSec": 1597.43,
        "decodeTokens": 9436,
        "peakRssMb": 36665,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T03:30:07.660Z",
        "trialId": "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u",
        "cellId": "r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1578,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 7,
          "Fix minimality": 8,
          "Honest verification": 7
        },
        "meanScore": 7.3,
        "justification": "Companion reproduction quotes the 80-versus-68 failure, validation quotes four green tests, and saved source has the minimal >= fix with a surviving threshold guard. The deterministic mechanical checks also pass. Verification is clear at unit level, though No unverified gaps is broader than the evidence and the notes could state that scope more carefully.",
        "artifactSha256": "bfa50a8bc64cd1eae87f9cf93547bea4207b2ca030e1b9901082222a74d02ac2",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 2
      },
      "startedAt": "2026-10-18T03:29:01.457Z",
      "finishedAt": "2026-10-18T03:29:26.516Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 25059,
        "timing": {
          "startedAt": "2026-10-18T03:29:01.457Z",
          "finishedAt": "2026-10-18T03:29:26.516Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 12582,
          "firstTurnTtftMs": 11022,
          "timeToFirstToolCallMs": 14142
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 118.26,
        "prefillTokensPerSec": 2393.5,
        "decodeTokens": 808,
        "peakRssMb": 29696,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T03:46:57.807Z",
        "trialId": "craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 2
      },
      "startedAt": "2026-10-18T03:29:35.370Z",
      "finishedAt": "2026-10-18T03:52:48.887Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1393517,
        "timing": {
          "startedAt": "2026-10-18T03:29:35.370Z",
          "finishedAt": "2026-10-18T03:52:48.887Z",
          "timeToFirstArtifactMs": 577424,
          "timeToLastArtifactWriteMs": 1191736,
          "timeToFirstTokenMs": 15696,
          "firstTurnTtftMs": 14052,
          "timeToFirstToolCallMs": 16040
        },
        "toolCalls": 129,
        "autoAnswers": 2,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 73.08,
        "prefillTokensPerSec": 1261.28,
        "decodeTokens": 58327,
        "peakRssMb": 38879,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T03:59:42.157Z",
        "trialId": "craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 2
      },
      "startedAt": "2026-10-18T03:52:57.982Z",
      "finishedAt": "2026-10-18T03:56:55.030Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 237048,
        "timing": {
          "startedAt": "2026-10-18T03:52:57.982Z",
          "finishedAt": "2026-10-18T03:56:55.030Z",
          "timeToFirstArtifactMs": 33119,
          "timeToLastArtifactWriteMs": 227812,
          "timeToFirstTokenMs": 18701,
          "firstTurnTtftMs": 12191,
          "timeToFirstToolCallMs": 19257
        },
        "toolCalls": 49,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 103.32,
        "prefillTokensPerSec": 1479.84,
        "decodeTokens": 12368,
        "peakRssMb": 38987,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T03:59:42.158Z",
        "trialId": "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 741,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 8,
          "numbering continuity": 8,
          "professionalism": 6,
          "completeness": 4,
          "print-polish": 8
        },
        "meanScore": 6.8,
        "justification": "Amounts total $5,415 and numbers 042–044 correctly exclude the old $1,200 bill; net-14/net-30 due dates are correct. Three invoices have clear tables and print margins. Bluestem's work date changes July 18 to September 18. Ledger.json still says last 041 with all three rows uninvoiced, so the books are not updated; 'went out' is unsupported.",
        "artifactSha256": "4a0826f10d12d72fc5f3b7b05932b5296f22bb9190e4476328c815947e364225",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 2
      },
      "startedAt": "2026-10-18T03:57:04.871Z",
      "finishedAt": "2026-10-18T04:09:41.902Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 757031,
        "timing": {
          "startedAt": "2026-10-18T03:57:04.871Z",
          "finishedAt": "2026-10-18T04:09:41.902Z",
          "timeToFirstArtifactMs": 237490,
          "timeToLastArtifactWriteMs": 724988,
          "timeToFirstTokenMs": 21223,
          "firstTurnTtftMs": 19137,
          "timeToFirstToolCallMs": 22967
        },
        "toolCalls": 76,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 74.49,
        "prefillTokensPerSec": 1124.48,
        "decodeTokens": 36715,
        "peakRssMb": 40396,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T04:14:25.069Z",
        "trialId": "craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 2
      },
      "startedAt": "2026-10-18T04:09:51.512Z",
      "finishedAt": "2026-10-18T04:15:46.128Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 354616,
        "timing": {
          "startedAt": "2026-10-18T04:09:51.512Z",
          "finishedAt": "2026-10-18T04:15:46.128Z",
          "timeToFirstArtifactMs": 92693,
          "timeToLastArtifactWriteMs": 322043,
          "timeToFirstTokenMs": 16674,
          "firstTurnTtftMs": 14570,
          "timeToFirstToolCallMs": 19194
        },
        "toolCalls": 47,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 88.52,
        "prefillTokensPerSec": 1417.58,
        "decodeTokens": 19477,
        "peakRssMb": 39155,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T04:29:25.583Z",
        "trialId": "craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 2
      },
      "startedAt": "2026-10-18T04:15:56.080Z",
      "finishedAt": "2026-10-18T04:41:37.471Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1541391,
        "timing": {
          "startedAt": "2026-10-18T04:15:56.080Z",
          "finishedAt": "2026-10-18T04:41:37.471Z",
          "timeToFirstArtifactMs": 213537,
          "timeToLastArtifactWriteMs": 1465687,
          "timeToFirstTokenMs": 21046,
          "firstTurnTtftMs": 19470,
          "timeToFirstToolCallMs": 26624
        },
        "toolCalls": 82,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 71.97,
        "prefillTokensPerSec": 1159.16,
        "decodeTokens": 47982,
        "peakRssMb": 37957,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T04:44:23.260Z",
        "trialId": "craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 2
      },
      "startedAt": "2026-10-18T04:41:46.672Z",
      "finishedAt": "2026-10-18T05:05:23.155Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-params:6:targetnone:fr1hkm6be:rp0:rf0\" stuck for 8m — 3 re-writes (39 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1416483,
        "timing": {
          "startedAt": "2026-10-18T04:41:46.672Z",
          "finishedAt": "2026-10-18T05:05:23.155Z",
          "timeToFirstArtifactMs": 316193,
          "timeToLastArtifactWriteMs": 1335759,
          "timeToFirstTokenMs": 21012,
          "firstTurnTtftMs": 18925,
          "timeToFirstToolCallMs": 23780
        },
        "toolCalls": 114,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 76.93,
        "prefillTokensPerSec": 1263.46,
        "decodeTokens": 48415,
        "peakRssMb": 38625,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T05:14:10.889Z",
        "trialId": "craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 2
      },
      "startedAt": "2026-10-18T05:05:32.673Z",
      "finishedAt": "2026-10-18T05:37:50.587Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a craftbook with a substantive embedded gate script, the report parses, and the task completed",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.1,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 7.5,
            "ruleId": "B_TWO_INTERVENTIONS",
            "summary": "No red flags or missing roles were observed; two interventions receive the midpoint between the fixed low/high-intervention anchors."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1937914,
        "timing": {
          "startedAt": "2026-10-18T05:05:32.673Z",
          "finishedAt": "2026-10-18T05:37:50.587Z",
          "timeToFirstArtifactMs": 416028,
          "timeToLastArtifactWriteMs": 1012227,
          "timeToFirstTokenMs": 20723,
          "firstTurnTtftMs": 19146,
          "timeToFirstToolCallMs": 24353
        },
        "toolCalls": 69,
        "autoAnswers": 2,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 70.46,
        "prefillTokensPerSec": 847.45,
        "decodeTokens": 38297,
        "peakRssMb": 40106,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T05:44:13.532Z",
        "trialId": "craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un",
      "cell": {
        "id": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 2
      },
      "startedAt": "2026-10-18T05:37:59.776Z",
      "finishedAt": "2026-10-18T06:06:23.839Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:5:targetnone:fr1tjwffw:rp0:rf0:m3\" stuck for 8m — 3 re-writes (18 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.8,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 62.5% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1704063,
        "timing": {
          "startedAt": "2026-10-18T05:37:59.776Z",
          "finishedAt": "2026-10-18T06:06:23.839Z",
          "timeToFirstArtifactMs": 632637,
          "timeToLastArtifactWriteMs": 844728,
          "timeToFirstTokenMs": 20697,
          "firstTurnTtftMs": 18609,
          "timeToFirstToolCallMs": 23276
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 62.31,
        "prefillTokensPerSec": 1069.53,
        "decodeTokens": 36281,
        "peakRssMb": 38413,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:04.762Z",
        "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un",
        "cellId": "r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:06:32.946Z",
      "finishedAt": "2026-10-18T06:07:12.339Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 39393,
        "timing": {
          "startedAt": "2026-10-18T06:06:32.946Z",
          "finishedAt": "2026-10-18T06:07:12.339Z",
          "timeToFirstArtifactMs": 35312,
          "timeToLastArtifactWriteMs": 35316,
          "timeToFirstTokenMs": 12763,
          "firstTurnTtftMs": 10872,
          "timeToFirstToolCallMs": 17818
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 115.13,
        "prefillTokensPerSec": 1738.48,
        "decodeTokens": 2099,
        "peakRssMb": 30146,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:48.646Z",
        "trialId": "tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 4146,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "Clear contrasting X/O colors, a winning-line glow and complete win/draw/reset handling make a usable small game. The implementation is compact and readable. Fixed 100px cells limit narrow-screen fit, initially empty buttons lack accessible names, and turn/result updates have no live announcement. Styling is coherent but basic.",
        "artifactSha256": "cc87c7ddede2c1e44ff9efb374f8d5bd2a17c124644d889b1fc8aa06645e3072",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:07:48.208Z",
      "finishedAt": "2026-10-18T06:08:52.478Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 64270,
        "timing": {
          "startedAt": "2026-10-18T06:07:48.208Z",
          "finishedAt": "2026-10-18T06:08:52.478Z",
          "timeToFirstArtifactMs": 59781,
          "timeToLastArtifactWriteMs": 59786,
          "timeToFirstTokenMs": 11863,
          "firstTurnTtftMs": 10314,
          "timeToFirstToolCallMs": 18417
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 135.48,
        "prefillTokensPerSec": 2081.43,
        "decodeTokens": 5395,
        "peakRssMb": 37932,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:48.648Z",
        "trialId": "petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/workspace/pet-shop-website/index.html",
        "artifactBytes": 13555,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 7,
          "functionalCompleteness": 6,
          "codeQuality": 6,
          "polish": 5
        },
        "meanScore": 6,
        "justification": "Warm consistent branding, a real PNG logo and product/cart interactions support the brief. Cart totals remain hidden because the total panel is never shown; the Contact link targets a missing id. Checkout and contact announce success without sending anything. Placeholder-only fields and an unstyled send button reduce polish.",
        "artifactSha256": "918086d30f17b5e14eb5d1a627cefaedcd16e6139a69c7ceccf9e7653a3aa31e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:09:28.480Z",
      "finishedAt": "2026-10-18T06:10:23.492Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 55012,
        "timing": {
          "startedAt": "2026-10-18T06:09:28.480Z",
          "finishedAt": "2026-10-18T06:10:23.492Z",
          "timeToFirstArtifactMs": 48745,
          "timeToLastArtifactWriteMs": 48749,
          "timeToFirstTokenMs": 11412,
          "firstTurnTtftMs": 9865,
          "timeToFirstToolCallMs": 15954
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 150.42,
        "prefillTokensPerSec": 2288.14,
        "decodeTokens": 4714,
        "peakRssMb": 30237,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:48.648Z",
        "trialId": "tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 11722,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 7,
          "codeQuality": 6,
          "polish": 6
        },
        "meanScore": 6.3,
        "justification": "Playable combat includes enemy fire, lives, waves, score, restart and explosions, with normalized diagonal movement and a resizing canvas. The drawn barrel points 90 degrees away from projectile travel. Movement and bullet speed depend on frame rate, controls are not explained beyond Enter, and losing a life has no invulnerability window.",
        "artifactSha256": "51856ed3edc4cf4f314430b9a14955952ebb474622f099c7bff3f1ee958cd7b4",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:10:59.651Z",
      "finishedAt": "2026-10-18T06:12:22.974Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 83323,
        "timing": {
          "startedAt": "2026-10-18T06:10:59.651Z",
          "finishedAt": "2026-10-18T06:12:22.974Z",
          "timeToFirstArtifactMs": 11848,
          "timeToLastArtifactWriteMs": 79700,
          "timeToFirstTokenMs": 8814,
          "firstTurnTtftMs": 7236,
          "timeToFirstToolCallMs": 11850
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 152.12,
        "prefillTokensPerSec": 2223.36,
        "decodeTokens": 3008,
        "peakRssMb": 35891,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:04.762Z",
        "trialId": "schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:12:32.028Z",
      "finishedAt": "2026-10-18T06:12:54.766Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 22738,
        "timing": {
          "startedAt": "2026-10-18T06:12:32.028Z",
          "finishedAt": "2026-10-18T06:12:54.766Z",
          "timeToFirstArtifactMs": 17673,
          "timeToLastArtifactWriteMs": 17675,
          "timeToFirstTokenMs": 9736,
          "firstTurnTtftMs": 8115,
          "timeToFirstToolCallMs": 10498
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 153.07,
        "prefillTokensPerSec": 2147.9,
        "decodeTokens": 785,
        "peakRssMb": 29399,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:04.762Z",
        "trialId": "failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:13:04.084Z",
      "finishedAt": "2026-10-18T06:13:31.319Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 27235,
        "timing": {
          "startedAt": "2026-10-18T06:13:04.084Z",
          "finishedAt": "2026-10-18T06:13:31.319Z",
          "timeToFirstArtifactMs": 25902,
          "timeToLastArtifactWriteMs": 25902,
          "timeToFirstTokenMs": 10163,
          "firstTurnTtftMs": 8205,
          "timeToFirstToolCallMs": 10669
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 116.47,
        "prefillTokensPerSec": 2148.21,
        "decodeTokens": 1740,
        "peakRssMb": 29241,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:14:04.762Z",
        "trialId": "symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:13:40.290Z",
      "finishedAt": "2026-10-18T06:15:42.808Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 122518,
        "timing": {
          "startedAt": "2026-10-18T06:13:40.290Z",
          "finishedAt": "2026-10-18T06:15:42.808Z",
          "timeToFirstArtifactMs": 24232,
          "timeToLastArtifactWriteMs": 116547,
          "timeToFirstTokenMs": 10002,
          "firstTurnTtftMs": 8424,
          "timeToFirstToolCallMs": 10856
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 102.97,
        "prefillTokensPerSec": 1589.75,
        "decodeTokens": 10672,
        "peakRssMb": 29913,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.568Z",
        "trialId": "data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:15:52.609Z",
      "finishedAt": "2026-10-18T06:19:16.301Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 203692,
        "timing": {
          "startedAt": "2026-10-18T06:15:52.609Z",
          "finishedAt": "2026-10-18T06:19:16.301Z",
          "timeToFirstArtifactMs": 200737,
          "timeToLastArtifactWriteMs": 200737,
          "timeToFirstTokenMs": 15378,
          "firstTurnTtftMs": 13748,
          "timeToFirstToolCallMs": 17943
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 84.25,
        "prefillTokensPerSec": 1350.9,
        "decodeTokens": 13559,
        "peakRssMb": 31872,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:19:25.296Z",
      "finishedAt": "2026-10-18T06:20:22.461Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 57165,
        "timing": {
          "startedAt": "2026-10-18T06:19:25.296Z",
          "finishedAt": "2026-10-18T06:20:22.461Z",
          "timeToFirstArtifactMs": 39386,
          "timeToLastArtifactWriteMs": 55896,
          "timeToFirstTokenMs": 15478,
          "firstTurnTtftMs": 13885,
          "timeToFirstToolCallMs": 16422
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 89.68,
        "prefillTokensPerSec": 1515.79,
        "decodeTokens": 2726,
        "peakRssMb": 31007,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:20:31.358Z",
      "finishedAt": "2026-10-18T06:21:03.411Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 32053,
        "timing": {
          "startedAt": "2026-10-18T06:20:31.358Z",
          "finishedAt": "2026-10-18T06:21:03.411Z",
          "timeToFirstArtifactMs": 27441,
          "timeToLastArtifactWriteMs": 27441,
          "timeToFirstTokenMs": 10774,
          "firstTurnTtftMs": 9186,
          "timeToFirstToolCallMs": 11491
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 87.46,
        "prefillTokensPerSec": 1573.71,
        "decodeTokens": 1416,
        "peakRssMb": 29245,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:21:12.317Z",
      "finishedAt": "2026-10-18T06:23:55.172Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on synthesis.md; the same scenario failure stayed unchanged. Last failure: - unsupported budget claim not present in the seeded sources: re-read finance.csv total_budget and line-item amounts, then replace every claimed amount with a source-backed value",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 162855,
        "timing": {
          "startedAt": "2026-10-18T06:21:12.317Z",
          "finishedAt": "2026-10-18T06:23:55.172Z",
          "timeToFirstArtifactMs": 68411,
          "timeToLastArtifactWriteMs": 158809,
          "timeToFirstTokenMs": 14965,
          "firstTurnTtftMs": 13380,
          "timeToFirstToolCallMs": 15948
        },
        "toolCalls": 47,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 95.56,
        "prefillTokensPerSec": 1500.89,
        "decodeTokens": 8421,
        "peakRssMb": 37845,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl",
        "cellId": "r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:24:04.886Z",
      "finishedAt": "2026-10-18T06:24:57.071Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 52185,
        "timing": {
          "startedAt": "2026-10-18T06:24:04.886Z",
          "finishedAt": "2026-10-18T06:24:57.071Z",
          "timeToFirstArtifactMs": 21467,
          "timeToLastArtifactWriteMs": 48701,
          "timeToFirstTokenMs": 9237,
          "firstTurnTtftMs": 7655,
          "timeToFirstToolCallMs": 11252
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 120.47,
        "prefillTokensPerSec": 1615.1,
        "decodeTokens": 2301,
        "peakRssMb": 32448,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:25:05.949Z",
      "finishedAt": "2026-10-18T06:27:08.409Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 122460,
        "timing": {
          "startedAt": "2026-10-18T06:25:05.949Z",
          "finishedAt": "2026-10-18T06:27:08.409Z",
          "timeToFirstArtifactMs": 64265,
          "timeToLastArtifactWriteMs": 119180,
          "timeToFirstTokenMs": 10801,
          "firstTurnTtftMs": 9207,
          "timeToFirstToolCallMs": 12042
        },
        "toolCalls": 15,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 111.46,
        "prefillTokensPerSec": 1395.98,
        "decodeTokens": 11488,
        "peakRssMb": 29882,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:27:17.238Z",
      "finishedAt": "2026-10-18T06:27:54.324Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37086,
        "timing": {
          "startedAt": "2026-10-18T06:27:17.238Z",
          "finishedAt": "2026-10-18T06:27:54.324Z",
          "timeToFirstArtifactMs": 32848,
          "timeToLastArtifactWriteMs": 35917,
          "timeToFirstTokenMs": 10678,
          "firstTurnTtftMs": 9086,
          "timeToFirstToolCallMs": 11336
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 86.88,
        "prefillTokensPerSec": 1510.66,
        "decodeTokens": 2061,
        "peakRssMb": 29549,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:29:08.569Z",
        "trialId": "plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:28:03.196Z",
      "finishedAt": "2026-10-18T06:29:55.916Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 112720,
        "timing": {
          "startedAt": "2026-10-18T06:28:03.196Z",
          "finishedAt": "2026-10-18T06:29:55.916Z",
          "timeToFirstArtifactMs": 20928,
          "timeToLastArtifactWriteMs": 109921,
          "timeToFirstTokenMs": 12536,
          "firstTurnTtftMs": 10952,
          "timeToFirstToolCallMs": 13391
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 131.96,
        "prefillTokensPerSec": 1723.56,
        "decodeTokens": 1198,
        "peakRssMb": 30540,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:46:15.475Z",
        "trialId": "meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1491,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 7,
          "grounding": 7,
          "actionability": 8,
          "brevity": 7
        },
        "meanScore": 7.3,
        "justification": "Clear decision, action, question and risk sections preserve the launch scope, owners, dates and field-map dependency; the CSV matches the four real actions. The final risk paragraph calls Jordan's review due August 30, although only scheduling it is due then. Repeated unresolved questions add some length, but the follow-up is readily executable.",
        "artifactSha256": "aad71b74c0465fbf1bda767fe68c5976ff8d6883e02c19e0144b69406dd26016",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:30:04.747Z",
      "finishedAt": "2026-10-18T06:32:55.401Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 170654,
        "timing": {
          "startedAt": "2026-10-18T06:30:04.747Z",
          "finishedAt": "2026-10-18T06:32:55.401Z",
          "timeToFirstArtifactMs": 32313,
          "timeToLastArtifactWriteMs": 85705,
          "timeToFirstTokenMs": 15638,
          "firstTurnTtftMs": 9112,
          "timeToFirstToolCallMs": 17627
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 119.58,
        "prefillTokensPerSec": 2108.33,
        "decodeTokens": 15038,
        "peakRssMb": 32347,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:46:15.477Z",
        "trialId": "craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 3211,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 5,
          "usefulness": 3
        },
        "meanScore": 4,
        "justification": "Scope separates drivers, formulas and scenarios, with live HTML calculations. AOV is claimed sourced as $100 although $4,760/40=$119. The 400-order sanity arithmetic is wrong; at 600 orders its formulas yield -$92.80, not +$387.20. The extra 60% factor discounts rates already described as preventable leakage, undermining the model.",
        "artifactSha256": "037b3677e033eae675aaee200133de538087d952c37b4275f765064be96206eb",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:33:04.814Z",
      "finishedAt": "2026-10-18T06:36:43.068Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 218254,
        "timing": {
          "startedAt": "2026-10-18T06:33:04.814Z",
          "finishedAt": "2026-10-18T06:36:43.068Z",
          "timeToFirstArtifactMs": 89883,
          "timeToLastArtifactWriteMs": 216900,
          "timeToFirstTokenMs": 15104,
          "firstTurnTtftMs": 13384,
          "timeToFirstToolCallMs": 16978
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 85.35,
        "prefillTokensPerSec": 1347.7,
        "decodeTokens": 9881,
        "peakRssMb": 37337,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:44:06.237Z",
        "trialId": "conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:36:52.779Z",
      "finishedAt": "2026-10-18T06:37:29.936Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 37157,
        "timing": {
          "startedAt": "2026-10-18T06:36:52.779Z",
          "finishedAt": "2026-10-18T06:37:29.936Z",
          "timeToFirstArtifactMs": 31322,
          "timeToLastArtifactWriteMs": 33880,
          "timeToFirstTokenMs": 12348,
          "firstTurnTtftMs": 10699,
          "timeToFirstToolCallMs": 12874
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 99.5,
        "prefillTokensPerSec": 1242.23,
        "decodeTokens": 1479,
        "peakRssMb": 32462,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:46:15.477Z",
        "trialId": "docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 834,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 2,
          "candor": 6,
          "usefulness": 6
        },
        "meanScore": 4.7,
        "justification": "Core theme tokens and page counts are correct, and Callout Emphasis is explicitly unresolved with a concrete recommendation. But the 48/32pt margins were never returned. Tightened spacing as the cause, no matching style, retained source formatting and all other content preserved are also unsupported claims stated as facts.",
        "artifactSha256": "5eb66ccc463cc13939117aae1e2f172b30aa15e4f4bf981e2e60409c32487b14",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:37:38.770Z",
      "finishedAt": "2026-10-18T06:39:51.474Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 132704,
        "timing": {
          "startedAt": "2026-10-18T06:37:38.770Z",
          "finishedAt": "2026-10-18T06:39:51.474Z",
          "timeToFirstArtifactMs": 39840,
          "timeToLastArtifactWriteMs": 128850,
          "timeToFirstTokenMs": 19914,
          "firstTurnTtftMs": 13307,
          "timeToFirstToolCallMs": 20384
        },
        "toolCalls": 28,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 90.98,
        "prefillTokensPerSec": 1707.75,
        "decodeTokens": 6573,
        "peakRssMb": 34415,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:46:15.477Z",
        "trialId": "craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1743,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 7,
          "outline-parity": 8,
          "grounding": 5,
          "convertibility": 8
        },
        "meanScore": 7,
        "justification": "Source packet cites the exact brief and all seven outline titles become matching H1s in order; metrics and the three required next actions are preserved in concise slides. The deck adds unsourced claims about lost trust, an untracked workflow, leadership demanding proof and enterprise rollout intent. Its packet also incorrectly says the explicitly stated hours were inferred.",
        "artifactSha256": "ed0f169639a6cc73bfbcbb2928bcc9a361852c1512e0d2fc7c0f87307962f503",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:40:01.281Z",
      "finishedAt": "2026-10-18T06:45:20.033Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: claim-calibration. Last failure: every mention of the “first programmer” label must be hedged in the same sentence as debated shorthand",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 318752,
        "timing": {
          "startedAt": "2026-10-18T06:40:01.281Z",
          "finishedAt": "2026-10-18T06:45:20.033Z",
          "timeToFirstArtifactMs": 101082,
          "timeToLastArtifactWriteMs": 313116,
          "timeToFirstTokenMs": 9807,
          "firstTurnTtftMs": 8171,
          "timeToFirstToolCallMs": 24864
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 94.87,
        "prefillTokensPerSec": 1388.8,
        "decodeTokens": 19495,
        "peakRssMb": 36686,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T06:46:42.585Z",
        "trialId": "wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 8449,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 3,
          "citationQuality": 5,
          "calibration": 4
        },
        "meanScore": 4,
        "justification": "Core chronology and nearby card ids are diluted by repetition and unsourced claims about Lovelace's motivations, corrections, algorithm mechanics and there being no prototype. Authorship is called disputed, then her work is described as the first published program without qualification. Interpretation is frequently presented as evidence, weakening grounding and calibration.",
        "artifactSha256": "ec8812baa48ba2a2aaf457bc2753f44ab4dbe0ee628cdf46ffe7b9cd25647d61",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 3
      },
      "startedAt": "2026-10-18T06:45:29.598Z",
      "finishedAt": "2026-10-18T07:35:12.645Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-week-plan passed 12 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2983047,
        "timing": {
          "startedAt": "2026-10-18T06:45:29.598Z",
          "finishedAt": "2026-10-18T07:35:12.645Z",
          "timeToFirstArtifactMs": 1346198,
          "timeToLastArtifactWriteMs": 2975922,
          "timeToFirstTokenMs": 22425,
          "firstTurnTtftMs": 15863,
          "timeToFirstToolCallMs": 24641
        },
        "toolCalls": 244,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 79.29,
        "prefillTokensPerSec": 1312.85,
        "decodeTokens": 181271,
        "peakRssMb": 44602,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T07:44:55.591Z",
        "trialId": "craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/workspace/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 3658,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 8,
          "prioritization": 3,
          "conflict-handling": 3,
          "realism": 4
        },
        "meanScore": 4.5,
        "justification": "All eight fetched events and three follow-ups are preserved. Moving Priya to 11:15 collides with a focus block; moving budget review to 10:30 retains the original clash. Budget and design preparation occur after their meetings, agenda-less meetings lack explicit agenda prep, and 'blocked in the calendar' claims an unrecorded side effect.",
        "artifactSha256": "042d85acc560a9d49697ba650dcafc4397c28474ee626431fc73586e72772e53",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:35:25.168Z",
      "finishedAt": "2026-10-18T07:40:40.508Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 315340,
        "timing": {
          "startedAt": "2026-10-18T07:35:25.168Z",
          "finishedAt": "2026-10-18T07:40:40.508Z",
          "timeToFirstArtifactMs": 37606,
          "timeToLastArtifactWriteMs": 310532,
          "timeToFirstTokenMs": 17760,
          "firstTurnTtftMs": 11192,
          "timeToFirstToolCallMs": 18298
        },
        "toolCalls": 50,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 98.94,
        "prefillTokensPerSec": 1819.89,
        "decodeTokens": 17651,
        "peakRssMb": 37952,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T07:44:03.815Z",
        "trialId": "craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2433,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:40:49.533Z",
      "finishedAt": "2026-10-18T07:43:32.280Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 162747,
        "timing": {
          "startedAt": "2026-10-18T07:40:49.533Z",
          "finishedAt": "2026-10-18T07:43:32.280Z",
          "timeToFirstArtifactMs": 28441,
          "timeToLastArtifactWriteMs": 130023,
          "timeToFirstTokenMs": 17474,
          "firstTurnTtftMs": 10997,
          "timeToFirstToolCallMs": 19065
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 95.2,
        "prefillTokensPerSec": 1719.74,
        "decodeTokens": 10683,
        "peakRssMb": 33378,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T07:44:55.593Z",
        "trialId": "craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 7607,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 4,
          "grounding": 3,
          "structure": 7,
          "actionability": 5
        },
        "meanScore": 4.8,
        "justification": "Six consistent APA-style entries supply all three annotations, with useful comparisons and preserved sample sizes. Allen’s 11% becomes an unsupported 11 percentage-point absolute reduction. Other ungrounded specifics include a cross-sectional design, non-random sampling, a phone/video-only hybrid and Williams proposing a checklist. Critiques often treat sparse cards as full papers.",
        "artifactSha256": "7b38e8777c5115fccea01b1830fc1726810bf8d9177709344da4e332d7a7dd3a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:43:42.072Z",
      "finishedAt": "2026-10-18T07:46:05.485Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 143413,
        "timing": {
          "startedAt": "2026-10-18T07:43:42.072Z",
          "finishedAt": "2026-10-18T07:46:05.485Z",
          "timeToFirstArtifactMs": 31522,
          "timeToLastArtifactWriteMs": 86608,
          "timeToFirstTokenMs": 17509,
          "firstTurnTtftMs": 10918,
          "timeToFirstToolCallMs": 19438
        },
        "toolCalls": 32,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 91.82,
        "prefillTokensPerSec": 1793.36,
        "decodeTokens": 6967,
        "peakRssMb": 34790,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:00:30.322Z",
        "trialId": "craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0",
        "cellId": "r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1630,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 5,
          "structure": 7,
          "convertibility": 8
        },
        "meanScore": 6.7,
        "justification": "Title, summary, metrics table and next actions form a compact, directly convertible document. Supplied figures and action names are preserved, including valid derived reductions. But it turns the top complaint into the single remaining complaint and adds customer frustration, repeat contacts and detailed causal benefits while claiming all facts derive from the brief.",
        "artifactSha256": "0507beb6a124b5572d15ea44c95eb9dfc1dc7e062e1ad5913fb1df500c63a7e7",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:46:14.455Z",
      "finishedAt": "2026-10-18T07:46:47.217Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 32762,
        "timing": {
          "startedAt": "2026-10-18T07:46:14.455Z",
          "finishedAt": "2026-10-18T07:46:47.217Z",
          "timeToFirstArtifactMs": 29948,
          "timeToLastArtifactWriteMs": 29978,
          "timeToFirstTokenMs": 10530,
          "firstTurnTtftMs": 8930,
          "timeToFirstToolCallMs": 11348
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 115.34,
        "prefillTokensPerSec": 1527.68,
        "decodeTokens": 1992,
        "peakRssMb": 29305,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T07:59:10.490Z",
        "trialId": "fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:46:56.384Z",
      "finishedAt": "2026-10-18T07:47:24.081Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 27697,
        "timing": {
          "startedAt": "2026-10-18T07:46:56.384Z",
          "finishedAt": "2026-10-18T07:47:24.081Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 15691,
          "firstTurnTtftMs": 14123,
          "timeToFirstToolCallMs": 17960
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.59,
        "prefillTokensPerSec": 1502.42,
        "decodeTokens": 743,
        "peakRssMb": 29655,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T07:59:10.490Z",
        "trialId": "dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:47:33.323Z",
      "finishedAt": "2026-10-18T07:48:15.377Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 7 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 42054,
        "timing": {
          "startedAt": "2026-10-18T07:47:33.323Z",
          "finishedAt": "2026-10-18T07:48:15.377Z",
          "timeToFirstArtifactMs": 36285,
          "timeToLastArtifactWriteMs": 39427,
          "timeToFirstTokenMs": 12305,
          "firstTurnTtftMs": 10763,
          "timeToFirstToolCallMs": 19687
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 104.82,
        "prefillTokensPerSec": 1904.82,
        "decodeTokens": 1920,
        "peakRssMb": 30391,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T07:59:10.490Z",
        "trialId": "interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:48:24.276Z",
      "finishedAt": "2026-10-18T07:50:17.769Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 113493,
        "timing": {
          "startedAt": "2026-10-18T07:48:24.276Z",
          "finishedAt": "2026-10-18T07:50:17.769Z",
          "timeToFirstArtifactMs": 84652,
          "timeToLastArtifactWriteMs": 84652,
          "timeToFirstTokenMs": 16849,
          "firstTurnTtftMs": 10318,
          "timeToFirstToolCallMs": 17440
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 105.83,
        "prefillTokensPerSec": 1936.35,
        "decodeTokens": 6649,
        "peakRssMb": 32100,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:00:30.324Z",
        "trialId": "craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3589,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 4,
          "usefulness": 5
        },
        "meanScore": 4.5,
        "justification": "Finds real SQL injection, logging, resource cleanup, masked errors and pagination defects. It nevertheless raises all three patterns the conventions explicitly forbid flagging: best-effort audit writes, == null and Express arity. The audit garbage-collection claim is unsound, several line citations drift, and the title says five new modules while the summary lists six.",
        "artifactSha256": "2b4da653e0c987e410a41b52463dde578600e3fbde5b635488e7d75a51e38cfe",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:50:27.578Z",
      "finishedAt": "2026-10-18T07:57:16.498Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 408920,
        "timing": {
          "startedAt": "2026-10-18T07:50:27.578Z",
          "finishedAt": "2026-10-18T07:57:16.498Z",
          "timeToFirstArtifactMs": 118633,
          "timeToLastArtifactWriteMs": 368294,
          "timeToFirstTokenMs": 17728,
          "firstTurnTtftMs": 11208,
          "timeToFirstToolCallMs": 19018
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 89.66,
        "prefillTokensPerSec": 1365.13,
        "decodeTokens": 29344,
        "peakRssMb": 34751,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:00:30.324Z",
        "trialId": "craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 12675,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 3,
          "usefulness": 4
        },
        "meanScore": 3.5,
        "justification": "Many spec fixes are concrete, including retryCount typing, errors and required fields. But the final report misses the absent history handler and undocumented GET /widgets/:id, calls distinct GET/POST paths a collision, and substitutes broad style advice for implementation reconciliation. Its recommendation references nonexistent F18 and misidentifies the blocking findings.",
        "artifactSha256": "3af381151ad701d34d23ebdb827f050cd0fd558e3a848c996bb0c75ff86d9725",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:57:26.125Z",
      "finishedAt": "2026-10-18T07:59:04.528Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 98403,
        "timing": {
          "startedAt": "2026-10-18T07:57:26.125Z",
          "finishedAt": "2026-10-18T07:59:04.528Z",
          "timeToFirstArtifactMs": 42196,
          "timeToLastArtifactWriteMs": 94771,
          "timeToFirstTokenMs": 13838,
          "firstTurnTtftMs": 12244,
          "timeToFirstToolCallMs": 15317
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 101.08,
        "prefillTokensPerSec": 1548.04,
        "decodeTokens": 4457,
        "peakRssMb": 32187,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:00:30.325Z",
        "trialId": "craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 4902,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 7,
          "systemic analysis": 6,
          "remediation": 5
        },
        "meanScore": 6,
        "justification": "Confirms both injections and the missing admin guard while preserving safe exceptions. Summary says three critical, but F3 is High. Systemic text wrongly says the other registrars accept requireAdmin; they accept requireUser. The execFile fix reads out.out from a stdout string, and full-database write impact is asserted without privilege evidence.",
        "artifactSha256": "6c6ad7c1815a1ddaf8a9ff7ad252602c6bde4fb240212336b75fee712e084553",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 3
      },
      "startedAt": "2026-10-18T07:59:13.469Z",
      "finishedAt": "2026-10-18T08:13:15.427Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 841958,
        "timing": {
          "startedAt": "2026-10-18T07:59:13.469Z",
          "finishedAt": "2026-10-18T08:13:15.427Z",
          "timeToFirstArtifactMs": 72807,
          "timeToLastArtifactWriteMs": 836301,
          "timeToFirstTokenMs": 16973,
          "firstTurnTtftMs": 15266,
          "timeToFirstToolCallMs": 19874
        },
        "toolCalls": 30,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 61.53,
        "prefillTokensPerSec": 791.81,
        "decodeTokens": 24521,
        "peakRssMb": 40762,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:14:07.676Z",
        "trialId": "large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:13:25.644Z",
      "finishedAt": "2026-10-18T08:17:55.129Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 269485,
        "timing": {
          "startedAt": "2026-10-18T08:13:25.644Z",
          "finishedAt": "2026-10-18T08:17:55.129Z",
          "timeToFirstArtifactMs": 66068,
          "timeToLastArtifactWriteMs": 266599,
          "timeToFirstTokenMs": 18463,
          "firstTurnTtftMs": 11913,
          "timeToFirstToolCallMs": 21529
        },
        "toolCalls": 73,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 99.95,
        "prefillTokensPerSec": 1576.37,
        "decodeTokens": 17729,
        "peakRssMb": 39279,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:29:42.519Z",
        "trialId": "craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 1668,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 7,
          "Invariant discipline": 8,
          "Reconciled verification": 7
        },
        "meanScore": 7.3,
        "justification": "All seven definition, re-export, import, call and log sites are enumerated, with both exclusions justified. Saved source shows the intended mechanical rename and 32 deterministic checks pass; verification quotes three passing tests. Counts reconcile clearly. The residual scan covers only the four edited files rather than demonstrating a full-tree post-change search.",
        "artifactSha256": "45cca06b5a7f85a155e1d86d8c697f902b68a1463aec9cfaf3aeead1cefc58d6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:18:04.628Z",
      "finishedAt": "2026-10-18T08:42:59.635Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-refactor-module:26:targetnone:frvmrl5y:rp0:rf0:m13\" stuck for 8m — 10 re-writes (24 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 86.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1495007,
        "timing": {
          "startedAt": "2026-10-18T08:18:04.628Z",
          "finishedAt": "2026-10-18T08:42:59.635Z",
          "timeToFirstArtifactMs": 27953,
          "timeToLastArtifactWriteMs": 1009181,
          "timeToFirstTokenMs": 15949,
          "firstTurnTtftMs": 9399,
          "timeToFirstToolCallMs": 27955
        },
        "toolCalls": 129,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.49,
        "prefillTokensPerSec": 1303.96,
        "decodeTokens": 103389,
        "peakRssMb": 41226,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:45:17.974Z",
        "trialId": "craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 4987,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 1,
          "Reversible-stage execution": 1,
          "Genuine deduplication": 0
        },
        "meanScore": 0.7,
        "justification": "Baseline describes buildReport and claims no duplicated helper, contradicting the seeded brief. Saved code removes reportLine/reportSummary and preserves repeated inner spaces. The harness still fails to import its original public API, while package.json changed. Identical green 28-test receipts across stages and 'no deviations' do not establish preserved behavior.",
        "artifactSha256": "9390d16fcd24e065aaea02ff96164a13d005f475712a88bccf51e3cf2e74281c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:43:21.325Z",
      "finishedAt": "2026-10-18T08:46:55.942Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 214617,
        "timing": {
          "startedAt": "2026-10-18T08:43:21.325Z",
          "finishedAt": "2026-10-18T08:46:55.942Z",
          "timeToFirstArtifactMs": 25298,
          "timeToLastArtifactWriteMs": 212859,
          "timeToFirstTokenMs": 18511,
          "firstTurnTtftMs": 11968,
          "timeToFirstToolCallMs": 19583
        },
        "toolCalls": 60,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 100.25,
        "prefillTokensPerSec": 1667.98,
        "decodeTokens": 11367,
        "peakRssMb": 39177,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:59:46.210Z",
        "trialId": "craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf",
        "cellId": "r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1802,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 7,
          "Fix minimality": 8,
          "Honest verification": 8
        },
        "meanScore": 7.7,
        "justification": "Reproduction quotes the real 80-versus-68 failure and identifies the added boundary test. Saved source makes only the correct >= threshold change; validation quotes four passing tests and the deterministic mutant-backed checks pass. Unlike a blanket completion claim, fix notes explicitly distinguish tested local behavior from an unverified downstream-consumer concern.",
        "artifactSha256": "d2f4467a39c96615b0dedac3366f8f14543155fb40cc17f6a76853c068e23254",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:47:05.030Z",
      "finishedAt": "2026-10-18T08:47:37.356Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 32326,
        "timing": {
          "startedAt": "2026-10-18T08:47:05.030Z",
          "finishedAt": "2026-10-18T08:47:37.356Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 15848,
          "firstTurnTtftMs": 14279,
          "timeToFirstToolCallMs": 17814
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.99,
        "prefillTokensPerSec": 1507.62,
        "decodeTokens": 709,
        "peakRssMb": 29439,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:59:07.263Z",
        "trialId": "craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:47:46.966Z",
      "finishedAt": "2026-10-18T08:53:10.451Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on task quarterly-brief/2 craftbook; missing signals stayed unchanged: task quarterly-brief/2 has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.. Last failure: task quarterly-brief/2 has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 323485,
        "timing": {
          "startedAt": "2026-10-18T08:47:46.966Z",
          "finishedAt": "2026-10-18T08:53:10.451Z",
          "timeToFirstArtifactMs": 146820,
          "timeToLastArtifactWriteMs": 320524,
          "timeToFirstTokenMs": 19581,
          "firstTurnTtftMs": 17935,
          "timeToFirstToolCallMs": 20929
        },
        "toolCalls": 44,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 86.41,
        "prefillTokensPerSec": 1380.36,
        "decodeTokens": 12569,
        "peakRssMb": 39396,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:59:07.263Z",
        "trialId": "craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:53:20.245Z",
      "finishedAt": "2026-10-18T08:57:42.336Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 262091,
        "timing": {
          "startedAt": "2026-10-18T08:53:20.245Z",
          "finishedAt": "2026-10-18T08:57:42.336Z",
          "timeToFirstArtifactMs": 38416,
          "timeToLastArtifactWriteMs": 254233,
          "timeToFirstTokenMs": 18185,
          "firstTurnTtftMs": 11665,
          "timeToFirstToolCallMs": 18705
        },
        "toolCalls": 53,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 104.89,
        "prefillTokensPerSec": 1512.47,
        "decodeTokens": 15335,
        "peakRssMb": 39131,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T08:59:46.211Z",
        "trialId": "craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 742,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 8,
          "numbering continuity": 8,
          "professionalism": 7,
          "completeness": 4,
          "print-polish": 8
        },
        "meanScore": 7,
        "justification": "Three invoices preserve ledger amounts, total $5,415, continue 042–044 and use correct net-14/net-30 due dates. Serif typography, clear totals and explicit print margins support paper delivery. Ondaatje is deliberately skipped. But ledger.json still has lastInvoiceNumber 041 and all three billable entries uninvoiced; Harbor’s invoice also omits its supplied billing contact.",
        "artifactSha256": "bfc1875a2738322c982b50f9af486cb8d9f29d9ddb7ff3db6cba09195f488022",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 3
      },
      "startedAt": "2026-10-18T08:57:52.158Z",
      "finishedAt": "2026-10-18T09:21:28.369Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1416211,
        "timing": {
          "startedAt": "2026-10-18T08:57:52.158Z",
          "finishedAt": "2026-10-18T09:21:28.369Z",
          "timeToFirstArtifactMs": 260633,
          "timeToLastArtifactWriteMs": 1413099,
          "timeToFirstTokenMs": 20748,
          "firstTurnTtftMs": 18670,
          "timeToFirstToolCallMs": 22627
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 80.15,
        "prefillTokensPerSec": 1182.74,
        "decodeTokens": 42194,
        "peakRssMb": 39967,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T09:29:17.466Z",
        "trialId": "craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 3
      },
      "startedAt": "2026-10-18T09:21:37.991Z",
      "finishedAt": "2026-10-18T09:28:28.398Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 410407,
        "timing": {
          "startedAt": "2026-10-18T09:21:37.991Z",
          "finishedAt": "2026-10-18T09:28:28.398Z",
          "timeToFirstArtifactMs": 157434,
          "timeToLastArtifactWriteMs": 404697,
          "timeToFirstTokenMs": 16468,
          "firstTurnTtftMs": 14391,
          "timeToFirstToolCallMs": 18754
        },
        "toolCalls": 36,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 75.82,
        "prefillTokensPerSec": 1325.16,
        "decodeTokens": 9951,
        "peakRssMb": 37254,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T09:29:17.467Z",
        "trialId": "craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 3
      },
      "startedAt": "2026-10-18T09:28:37.562Z",
      "finishedAt": "2026-10-18T09:42:06.102Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 808540,
        "timing": {
          "startedAt": "2026-10-18T09:28:37.562Z",
          "finishedAt": "2026-10-18T09:42:06.102Z",
          "timeToFirstArtifactMs": 203161,
          "timeToLastArtifactWriteMs": 801409,
          "timeToFirstTokenMs": 20527,
          "firstTurnTtftMs": 18952,
          "timeToFirstToolCallMs": 22845
        },
        "toolCalls": 60,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 73.02,
        "prefillTokensPerSec": 1246.15,
        "decodeTokens": 23471,
        "peakRssMb": 38651,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T09:44:20.578Z",
        "trialId": "craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 3
      },
      "startedAt": "2026-10-18T09:42:15.849Z",
      "finishedAt": "2026-10-18T10:09:12.613Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-params:6:targetnone:frp4roae:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 35 tool calls (4 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1616764,
        "timing": {
          "startedAt": "2026-10-18T09:42:15.849Z",
          "finishedAt": "2026-10-18T10:09:12.613Z",
          "timeToFirstArtifactMs": 533194,
          "timeToLastArtifactWriteMs": 1549582,
          "timeToFirstTokenMs": 20612,
          "firstTurnTtftMs": 18533,
          "timeToFirstToolCallMs": 22867
        },
        "toolCalls": 72,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 67.43,
        "prefillTokensPerSec": 1117.44,
        "decodeTokens": 28461,
        "peakRssMb": 35563,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T10:14:17.985Z",
        "trialId": "craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 3
      },
      "startedAt": "2026-10-18T10:09:21.592Z",
      "finishedAt": "2026-10-18T10:28:41.284Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-gate-script:3:targetnone:fr1dbmw0o:rp0:rf0\" stuck for 13m — 3 re-writes (30 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.7,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 60% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1159692,
        "timing": {
          "startedAt": "2026-10-18T10:09:21.592Z",
          "finishedAt": "2026-10-18T10:28:41.284Z",
          "timeToFirstArtifactMs": 344683,
          "timeToLastArtifactWriteMs": 1155797,
          "timeToFirstTokenMs": 20130,
          "firstTurnTtftMs": 18419,
          "timeToFirstToolCallMs": 22915
        },
        "toolCalls": 51,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 80.03,
        "prefillTokensPerSec": 1053.36,
        "decodeTokens": 26889,
        "peakRssMb": 40245,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T10:29:16.965Z",
        "trialId": "craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4",
      "cell": {
        "id": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "qwen3.6-35b-a3b-q4",
        "targetId": "qwen3.6-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 3
      },
      "startedAt": "2026-10-18T10:28:50.605Z",
      "finishedAt": "2026-10-18T10:52:16.166Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:6:targetnone:fr1tjwffw:rp0:rf0:m3\" stuck for 8m — 9 re-writes (30 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "infra",
        "includedInFixedModelAggregate": false,
        "failureClassRule": "gate-unevaluable"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1405561,
        "timing": {
          "startedAt": "2026-10-18T10:28:50.605Z",
          "finishedAt": "2026-10-18T10:52:16.166Z",
          "timeToFirstArtifactMs": 366536,
          "timeToLastArtifactWriteMs": 1247954,
          "timeToFirstTokenMs": 20681,
          "firstTurnTtftMs": 18578,
          "timeToFirstToolCallMs": 22864
        },
        "toolCalls": 95,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.6-35b-a3b-q4"
        },
        "decodeTokensPerSec": 69.69,
        "prefillTokensPerSec": 1114.76,
        "decodeTokens": 36755,
        "peakRssMb": 42104,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:04:06.625Z",
        "trialId": "craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4",
        "cellId": "r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "qwen3.6-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:04:57.118Z",
      "finishedAt": "2026-10-18T11:06:57.961Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 120843,
        "timing": {
          "startedAt": "2026-10-18T11:04:57.118Z",
          "finishedAt": "2026-10-18T11:06:57.961Z",
          "timeToFirstArtifactMs": 113942,
          "timeToLastArtifactWriteMs": 113947,
          "timeToFirstTokenMs": 20191,
          "firstTurnTtftMs": 18312,
          "timeToFirstToolCallMs": 31345
        },
        "toolCalls": 4,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 33.3,
        "prefillTokensPerSec": 530.06,
        "decodeTokens": 2719,
        "peakRssMb": 41748,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:14:52.163Z",
        "trialId": "tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs",
        "cellId": "r1-qwen3.8-27b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 3601,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "Coherent dark board with contrasting marks, win highlighting, draw detection, turn display and reset meets the brief. Small, readable code guards occupied cells and finished games. Fixed 100px cells limit small-screen fit; empty buttons have no accessible labels and status changes lack live announcements. Presentation is basic.",
        "artifactSha256": "92b4f7b6f3e55fef2e17abff8d4b58db2160651ea77e1fa6dbbd6d92a8eb16b1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:07:32.718Z",
      "finishedAt": "2026-10-18T11:11:22.983Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 230265,
        "timing": {
          "startedAt": "2026-10-18T11:07:32.718Z",
          "finishedAt": "2026-10-18T11:11:22.983Z",
          "timeToFirstArtifactMs": 200424,
          "timeToLastArtifactWriteMs": 200429,
          "timeToFirstTokenMs": 19595,
          "firstTurnTtftMs": 18046,
          "timeToFirstToolCallMs": 49765
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.25,
        "prefillTokensPerSec": 446.16,
        "decodeTokens": 4819,
        "peakRssMb": 50362,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:14:52.165Z",
        "trialId": "petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl",
        "cellId": "r1-qwen3.8-27b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/workspace/pet-shop-website/index.html",
        "artifactBytes": 6678,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 7,
          "codeQuality": 6,
          "polish": 5
        },
        "meanScore": 6,
        "justification": "Coherent green/cream sections and a generated PNG logo meet the site brief. Mobile navigation lacks wrapping, form fields lack labels, and contact submission promises a reply while only clearing the form; nothing is delivered. Static product cards and generic copy offer little polish.",
        "artifactSha256": "006dc60539001204e1470d1bd5ffbd44ea8b1eee7467b42e19a32229c449a6e3",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:12:00.664Z",
      "finishedAt": "2026-10-18T11:20:48.325Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 527661,
        "timing": {
          "startedAt": "2026-10-18T11:12:00.664Z",
          "finishedAt": "2026-10-18T11:20:48.325Z",
          "timeToFirstArtifactMs": 288089,
          "timeToLastArtifactWriteMs": 518483,
          "timeToFirstTokenMs": 25903,
          "firstTurnTtftMs": 24353,
          "timeToFirstToolCallMs": 77418
        },
        "toolCalls": 13,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.73,
        "prefillTokensPerSec": 391.76,
        "decodeTokens": 8981,
        "peakRssMb": 45520,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:29:31.724Z",
        "trialId": "tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a",
        "cellId": "r1-qwen3.8-27b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 7728,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 6,
          "polish": 5
        },
        "meanScore": 6.3,
        "justification": "Single-file game includes movement, aligned aiming, player/enemy fire, collisions, score, lives and restart with observable input. Basic tanks and grid are readable. Movement and cooldowns depend on frame rate; arrow keys can scroll, focus loss leaves keys held, and the fixed canvas has no controls guide, hit feedback or responsive fit.",
        "artifactSha256": "6f6f35cf910c5f8f3da4ba590940a40a41b28d5ba67a96d67a961f6fb790a457",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:21:23.872Z",
      "finishedAt": "2026-10-18T11:25:48.035Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 264163,
        "timing": {
          "startedAt": "2026-10-18T11:21:23.872Z",
          "finishedAt": "2026-10-18T11:25:48.035Z",
          "timeToFirstArtifactMs": 22463,
          "timeToLastArtifactWriteMs": 258745,
          "timeToFirstTokenMs": 13589,
          "firstTurnTtftMs": 12001,
          "timeToFirstToolCallMs": 22465
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 31.84,
        "prefillTokensPerSec": 472.25,
        "decodeTokens": 3159,
        "peakRssMb": 47847,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:29:04.944Z",
        "trialId": "schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z",
        "cellId": "r1-qwen3.8-27b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:25:56.835Z",
      "finishedAt": "2026-10-18T11:26:54.948Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 58113,
        "timing": {
          "startedAt": "2026-10-18T11:25:56.835Z",
          "finishedAt": "2026-10-18T11:26:54.948Z",
          "timeToFirstArtifactMs": 52406,
          "timeToLastArtifactWriteMs": 52408,
          "timeToFirstTokenMs": 18631,
          "firstTurnTtftMs": 17004,
          "timeToFirstToolCallMs": 20185
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 29.4,
        "prefillTokensPerSec": 374.17,
        "decodeTokens": 620,
        "peakRssMb": 37696,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:29:04.944Z",
        "trialId": "failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj",
        "cellId": "r1-qwen3.8-27b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:27:03.843Z",
      "finishedAt": "2026-10-18T11:28:16.439Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 72596,
        "timing": {
          "startedAt": "2026-10-18T11:27:03.843Z",
          "finishedAt": "2026-10-18T11:28:16.439Z",
          "timeToFirstArtifactMs": 67202,
          "timeToLastArtifactWriteMs": 67202,
          "timeToFirstTokenMs": 18199,
          "firstTurnTtftMs": 16604,
          "timeToFirstToolCallMs": 20245
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.49,
        "prefillTokensPerSec": 400.6,
        "decodeTokens": 932,
        "peakRssMb": 37305,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:29:04.944Z",
        "trialId": "symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50",
        "cellId": "r1-qwen3.8-27b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:28:25.943Z",
      "finishedAt": "2026-10-18T11:29:53.664Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 87721,
        "timing": {
          "startedAt": "2026-10-18T11:28:25.943Z",
          "finishedAt": "2026-10-18T11:29:53.664Z",
          "timeToFirstArtifactMs": 68841,
          "timeToLastArtifactWriteMs": 68841,
          "timeToFirstTokenMs": 18076,
          "firstTurnTtftMs": 16486,
          "timeToFirstToolCallMs": 20467
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.65,
        "prefillTokensPerSec": 406.43,
        "decodeTokens": 1520,
        "peakRssMb": 38122,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:44:04.865Z",
        "trialId": "data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0",
        "cellId": "r1-qwen3.8-27b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:30:03.109Z",
      "finishedAt": "2026-10-18T11:51:46.333Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1303224,
        "timing": {
          "startedAt": "2026-10-18T11:30:03.109Z",
          "finishedAt": "2026-10-18T11:51:46.333Z",
          "timeToFirstArtifactMs": 1177805,
          "timeToLastArtifactWriteMs": 1296964,
          "timeToFirstTokenMs": 34150,
          "firstTurnTtftMs": 32551,
          "timeToFirstToolCallMs": 39160
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.33,
        "prefillTokensPerSec": 368.91,
        "decodeTokens": 23652,
        "peakRssMb": 47743,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:59:04.701Z",
        "trialId": "incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo",
        "cellId": "r1-qwen3.8-27b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:51:55.486Z",
      "finishedAt": "2026-10-18T11:54:43.675Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 168189,
        "timing": {
          "startedAt": "2026-10-18T11:51:55.486Z",
          "finishedAt": "2026-10-18T11:54:43.675Z",
          "timeToFirstArtifactMs": 86596,
          "timeToLastArtifactWriteMs": 161787,
          "timeToFirstTokenMs": 34678,
          "firstTurnTtftMs": 33078,
          "timeToFirstToolCallMs": 40131
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.96,
        "prefillTokensPerSec": 413.34,
        "decodeTokens": 2809,
        "peakRssMb": 40150,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:59:04.702Z",
        "trialId": "ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e",
        "cellId": "r1-qwen3.8-27b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:54:52.957Z",
      "finishedAt": "2026-10-18T11:57:15.909Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 142952,
        "timing": {
          "startedAt": "2026-10-18T11:54:52.957Z",
          "finishedAt": "2026-10-18T11:57:15.909Z",
          "timeToFirstArtifactMs": 140269,
          "timeToLastArtifactWriteMs": 140269,
          "timeToFirstTokenMs": 18524,
          "firstTurnTtftMs": 16937,
          "timeToFirstToolCallMs": 140270
        },
        "toolCalls": 1,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.87,
        "prefillTokensPerSec": 426.19,
        "decodeTokens": 2175,
        "peakRssMb": 36814,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T11:59:04.702Z",
        "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs",
        "cellId": "r1-qwen3.8-27b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-18T11:57:25.352Z",
      "finishedAt": "2026-10-18T12:04:09.368Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 404016,
        "timing": {
          "startedAt": "2026-10-18T11:57:25.352Z",
          "finishedAt": "2026-10-18T12:04:09.368Z",
          "timeToFirstArtifactMs": 191321,
          "timeToLastArtifactWriteMs": 399871,
          "timeToFirstTokenMs": 34515,
          "firstTurnTtftMs": 32925,
          "timeToFirstToolCallMs": 37642
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.84,
        "prefillTokensPerSec": 403.21,
        "decodeTokens": 5391,
        "peakRssMb": 48162,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:14:07.758Z",
        "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9",
        "cellId": "r1-qwen3.8-27b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:04:18.855Z",
      "finishedAt": "2026-10-18T12:05:21.350Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 62495,
        "timing": {
          "startedAt": "2026-10-18T12:04:18.855Z",
          "finishedAt": "2026-10-18T12:05:21.350Z",
          "timeToFirstArtifactMs": 22568,
          "timeToLastArtifactWriteMs": 59769,
          "timeToFirstTokenMs": 13642,
          "firstTurnTtftMs": 12049,
          "timeToFirstToolCallMs": 22570
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.7,
        "prefillTokensPerSec": 452.45,
        "decodeTokens": 624,
        "peakRssMb": 38589,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:14:07.759Z",
        "trialId": "constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu",
        "cellId": "r1-qwen3.8-27b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:05:30.895Z",
      "finishedAt": "2026-10-18T12:07:03.629Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 92734,
        "timing": {
          "startedAt": "2026-10-18T12:05:30.895Z",
          "finishedAt": "2026-10-18T12:07:03.629Z",
          "timeToFirstArtifactMs": 88681,
          "timeToLastArtifactWriteMs": 88681,
          "timeToFirstTokenMs": 18314,
          "firstTurnTtftMs": 16733,
          "timeToFirstToolCallMs": 19947
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 31.25,
        "prefillTokensPerSec": 423.3,
        "decodeTokens": 2096,
        "peakRssMb": 37678,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:14:07.759Z",
        "trialId": "records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop",
        "cellId": "r1-qwen3.8-27b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:07:13.025Z",
      "finishedAt": "2026-10-18T12:09:15.728Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 122703,
        "timing": {
          "startedAt": "2026-10-18T12:07:13.025Z",
          "finishedAt": "2026-10-18T12:09:15.728Z",
          "timeToFirstArtifactMs": 117617,
          "timeToLastArtifactWriteMs": 117617,
          "timeToFirstTokenMs": 18072,
          "firstTurnTtftMs": 16481,
          "timeToFirstToolCallMs": 117619
        },
        "toolCalls": 1,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 18.99,
        "prefillTokensPerSec": 446.75,
        "decodeTokens": 1890,
        "peakRssMb": 36812,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:14:07.759Z",
        "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi",
        "cellId": "r1-qwen3.8-27b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:09:25.276Z",
      "finishedAt": "2026-10-18T12:11:58.679Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 153403,
        "timing": {
          "startedAt": "2026-10-18T12:09:25.276Z",
          "finishedAt": "2026-10-18T12:11:58.679Z",
          "timeToFirstArtifactMs": 48762,
          "timeToLastArtifactWriteMs": 147854,
          "timeToFirstTokenMs": 25193,
          "firstTurnTtftMs": 23610,
          "timeToFirstToolCallMs": 27365
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 35.66,
        "prefillTokensPerSec": 453,
        "decodeTokens": 890,
        "peakRssMb": 38773,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:14:34.655Z",
        "trialId": "meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd",
        "cellId": "r1-qwen3.8-27b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1469,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 8,
          "grounding": 7,
          "actionability": 8,
          "brevity": 8
        },
        "meanScore": 7.8,
        "justification": "Clear decisions, four dated actions, the field-map dependency, unscheduled legal review and unowned rollback communications make the brief executable and concise. It correctly rejects stale proposals. The open-question text says Jordan will run the review by August 30, while the transcript only commits to scheduling it by then; the action table is correct.",
        "artifactSha256": "08a0cbb532dc826214e0d8b81c587fd886ccfa5bf617b3c2340ab62ea42a844f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:12:08.032Z",
      "finishedAt": "2026-10-18T12:27:52.279Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 944247,
        "timing": {
          "startedAt": "2026-10-18T12:12:08.032Z",
          "finishedAt": "2026-10-18T12:27:52.279Z",
          "timeToFirstArtifactMs": 87994,
          "timeToLastArtifactWriteMs": 581373,
          "timeToFirstTokenMs": 25379,
          "firstTurnTtftMs": 18873,
          "timeToFirstToolCallMs": 40522
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.3,
        "prefillTokensPerSec": 423.28,
        "decodeTokens": 18807,
        "peakRssMb": 46077,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:29:48.272Z",
        "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio",
        "cellId": "r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 3410,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 7
        },
        "meanScore": 7.5,
        "justification": "Scope defines the question, primary output, six editable drivers, monthly formulas, reconciliation checks and a correctly hand-calculated baseline before implementation. Defaults are explicitly separate from the seeded records, but their business basis is unexplained. It lacks a changed-input sanity case and leaves break-even as fractional units rather than the minimum whole units needed.",
        "artifactSha256": "b24882f0899c39d2ca2d486605239a7d1753dfeedcadd387f5e4afbbd305235f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:28:01.865Z",
      "finishedAt": "2026-10-18T12:44:58.676Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1016811,
        "timing": {
          "startedAt": "2026-10-18T12:28:01.865Z",
          "finishedAt": "2026-10-18T12:44:58.676Z",
          "timeToFirstArtifactMs": 430696,
          "timeToLastArtifactWriteMs": 1012894,
          "timeToFirstTokenMs": 34824,
          "firstTurnTtftMs": 33219,
          "timeToFirstToolCallMs": 38257
        },
        "toolCalls": 39,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.74,
        "prefillTokensPerSec": 404.62,
        "decodeTokens": 11825,
        "peakRssMb": 48887,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T12:59:06.350Z",
        "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1",
        "cellId": "r1-qwen3.8-27b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:45:08.024Z",
      "finishedAt": "2026-10-18T12:46:40.984Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 92960,
        "timing": {
          "startedAt": "2026-10-18T12:45:08.024Z",
          "finishedAt": "2026-10-18T12:46:40.984Z",
          "timeToFirstArtifactMs": 89037,
          "timeToLastArtifactWriteMs": 89037,
          "timeToFirstTokenMs": 22812,
          "firstTurnTtftMs": 21152,
          "timeToFirstToolCallMs": 23499
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 32.26,
        "prefillTokensPerSec": 384.66,
        "decodeTokens": 1566,
        "peakRssMb": 44772,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T13:00:02.188Z",
        "trialId": "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a",
        "cellId": "r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1064,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 3,
          "candor": 6,
          "usefulness": 6
        },
        "meanScore": 5,
        "justification": "Reports required theme values, page counts and unresolved style. It invents tighter spacing as the shorter deck’s cause, says no matching style existed and original formatting remained, and asserts contentPreserved:true despite these being absent from the supplied tool facts. An explicit emphasis override is actionable; keep-as-is advice rests on unsupported assurances.",
        "artifactSha256": "acdff7460e2374b8af18b945833833adbac7f230c18bbdcef76cb4e126759326",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:46:50.237Z",
      "finishedAt": "2026-10-18T12:58:36.052Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 705815,
        "timing": {
          "startedAt": "2026-10-18T12:46:50.237Z",
          "finishedAt": "2026-10-18T12:58:36.052Z",
          "timeToFirstArtifactMs": 180443,
          "timeToLastArtifactWriteMs": 699839,
          "timeToFirstTokenMs": 37453,
          "firstTurnTtftMs": 30856,
          "timeToFirstToolCallMs": 41681
        },
        "toolCalls": 38,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.66,
        "prefillTokensPerSec": 391.63,
        "decodeTokens": 12787,
        "peakRssMb": 49793,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T13:00:02.189Z",
        "trialId": "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z",
        "cellId": "r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 2121,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 8,
          "outline-parity": 8,
          "grounding": 7,
          "convertibility": 8
        },
        "meanScore": 7.8,
        "justification": "Cites the exact brief, preserves seven outline titles once and in order, and uses concise H1 slide blocks. Metrics and named actions are correct; opening framing is marked as a proposal. Claims that the actions close the gap and training reduces repeat handling errors add causal certainty beyond the brief. Some source-policing copy belongs outside the slides.",
        "artifactSha256": "3b7c3f755c977abfbffa3ba34d2587310d5a828ddd76c9dddfc3de97d77c9aaf",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 1
      },
      "startedAt": "2026-10-18T12:58:46.873Z",
      "finishedAt": "2026-10-18T13:05:59.519Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 432646,
        "timing": {
          "startedAt": "2026-10-18T12:58:46.873Z",
          "finishedAt": "2026-10-18T13:05:59.519Z",
          "timeToFirstArtifactMs": 423221,
          "timeToLastArtifactWriteMs": 423221,
          "timeToFirstTokenMs": 15354,
          "firstTurnTtftMs": 13714,
          "timeToFirstToolCallMs": 99171
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.78,
        "prefillTokensPerSec": 404.89,
        "decodeTokens": 8788,
        "peakRssMb": 38514,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T13:14:59.434Z",
        "trialId": "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o",
        "cellId": "r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5593,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 6,
          "grounding": 6,
          "citationQuality": 7,
          "calibration": 8
        },
        "meanScore": 6.8,
        "justification": "Preserves chronology, cites the three relevant cards beside claims, and consistently treats first programmer as disputed shorthand. Much of the brief repeats the extracts instead of integrating them. Claims that the Notes are the main surviving record and that later computing inherited specific design features exceed the cards, despite attached citations.",
        "artifactSha256": "c207370b1adc3e8f0b4d0a50b3cfdca13a51e554990d2e29e038814e29996389",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 1
      },
      "startedAt": "2026-10-18T13:06:08.438Z",
      "finishedAt": "2026-10-18T14:31:14.981Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "chat stalled for 302s — daemon reachable but issuing no model turns (soft digest stuck at 0562b69363ded3b0) (re-engage nudge was sent and ignored)",
        "failureMode": "chat-stalled",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 1.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 5106543,
        "timing": {
          "startedAt": "2026-10-18T13:06:08.438Z",
          "finishedAt": "2026-10-18T14:31:14.981Z",
          "timeToFirstArtifactMs": 276682,
          "timeToLastArtifactWriteMs": 3129434,
          "timeToFirstTokenMs": 47583,
          "firstTurnTtftMs": 41053,
          "timeToFirstToolCallMs": 51944
        },
        "toolCalls": 114,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.73,
        "prefillTokensPerSec": 371.56,
        "decodeTokens": 76269,
        "peakRssMb": 51616,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T14:45:34.196Z",
        "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj",
        "cellId": "r1-qwen3.8-27b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 8507,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 1,
          "prioritization": 2,
          "conflict-handling": 0,
          "realism": 1
        },
        "meanScore": 1,
        "justification": "Admits no calendar fetch, then invents a nine-meeting October week while claiming eight; the fixture is July 13–17 and its budget/Priya overlap is absent. Mock calendar requests are empty. Wednesday focus overlaps prep, Monday SOW work is consumed by meeting prep, and the fallback Tuesday move collides with Team sync. Follow-up text is retained, but the plan is not calendar-grounded.",
        "artifactSha256": "4f76264a16ff38e4d31e6fe3b4abadcec2453c0e747a5f41082746faf5d2b6c3",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 1
      },
      "startedAt": "2026-10-18T14:31:25.796Z",
      "finishedAt": "2026-10-18T14:51:19.226Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1193430,
        "timing": {
          "startedAt": "2026-10-18T14:31:25.796Z",
          "finishedAt": "2026-10-18T14:51:19.226Z",
          "timeToFirstArtifactMs": 139312,
          "timeToLastArtifactWriteMs": 1188309,
          "timeToFirstTokenMs": 33921,
          "firstTurnTtftMs": 27416,
          "timeToFirstToolCallMs": 37812
        },
        "toolCalls": 42,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 26.57,
        "prefillTokensPerSec": 447.38,
        "decodeTokens": 22053,
        "peakRssMb": 50086,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T14:59:07.826Z",
        "trialId": "craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4",
        "cellId": "r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 3953,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 1
      },
      "startedAt": "2026-10-18T14:51:30.330Z",
      "finishedAt": "2026-10-18T14:58:16.863Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 406533,
        "timing": {
          "startedAt": "2026-10-18T14:51:30.330Z",
          "finishedAt": "2026-10-18T14:58:16.863Z",
          "timeToFirstArtifactMs": 72062,
          "timeToLastArtifactWriteMs": 291596,
          "timeToFirstTokenMs": 27610,
          "firstTurnTtftMs": 21053,
          "timeToFirstToolCallMs": 29767
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.53,
        "prefillTokensPerSec": 427.41,
        "decodeTokens": 6280,
        "peakRssMb": 46353,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T14:59:48.135Z",
        "trialId": "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d",
        "cellId": "r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 5766,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 7,
          "grounding": 5,
          "structure": 8,
          "actionability": 6
        },
        "meanScore": 6.5,
        "justification": "All six seeded sources and key figures are retained with uniform, scannable annotations and useful caveats about missing metadata. The cards do not supply Allen’s usual-care comparator, Patel’s in-person-only cost-parity comparison, or a checklist produced by Williams; regulatory/platform-change claims are also unsupported. Relevance is clear but recommendations remain broad.",
        "artifactSha256": "91cda0ae408139446645cab1bf76ea57d3deadd1a68a8a1798ec7123a86a95d1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 1
      },
      "startedAt": "2026-10-18T14:58:25.853Z",
      "finishedAt": "2026-10-18T15:05:02.787Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 396934,
        "timing": {
          "startedAt": "2026-10-18T14:58:25.853Z",
          "finishedAt": "2026-10-18T15:05:02.787Z",
          "timeToFirstArtifactMs": 79541,
          "timeToLastArtifactWriteMs": 221587,
          "timeToFirstTokenMs": 31145,
          "firstTurnTtftMs": 24536,
          "timeToFirstToolCallMs": 40048
        },
        "toolCalls": 35,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.55,
        "prefillTokensPerSec": 409.63,
        "decodeTokens": 5397,
        "peakRssMb": 48468,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T15:14:47.965Z",
        "trialId": "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7",
        "cellId": "r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1629,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 6,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 7.3,
        "justification": "Correct metrics and all three named actions appear in a concise, complete Markdown source with summary, table and clear sections. It overstates the top complaint as the only unresolved friction, supplies an unsupported interpretation of refund leakage, and promises the proposed emails/training will end specific problems. Structure is ready for direct DOCX conversion.",
        "artifactSha256": "376e537765ab690e1e4e347b5b0bcdb63c23f2af3b191f01f04252c3b6167da5",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:05:14.345Z",
      "finishedAt": "2026-10-18T15:07:27.542Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 133197,
        "timing": {
          "startedAt": "2026-10-18T15:05:14.345Z",
          "finishedAt": "2026-10-18T15:07:27.542Z",
          "timeToFirstArtifactMs": 81555,
          "timeToLastArtifactWriteMs": 128907,
          "timeToFirstTokenMs": 17487,
          "firstTurnTtftMs": 15880,
          "timeToFirstToolCallMs": 20659
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 30.55,
        "prefillTokensPerSec": 398.68,
        "decodeTokens": 3028,
        "peakRssMb": 38048,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T15:14:14.792Z",
        "trialId": "fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr",
        "cellId": "r1-qwen3.8-27b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:07:36.652Z",
      "finishedAt": "2026-10-18T15:08:54.749Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 78097,
        "timing": {
          "startedAt": "2026-10-18T15:07:36.652Z",
          "finishedAt": "2026-10-18T15:08:54.749Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 36571,
          "firstTurnTtftMs": 34955,
          "timeToFirstToolCallMs": 46389
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.3,
        "prefillTokensPerSec": 417.98,
        "decodeTokens": 737,
        "peakRssMb": 37541,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T15:14:14.792Z",
        "trialId": "dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav",
        "cellId": "r1-qwen3.8-27b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:09:03.717Z",
      "finishedAt": "2026-10-18T15:14:37.980Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 5 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 334263,
        "timing": {
          "startedAt": "2026-10-18T15:09:03.717Z",
          "finishedAt": "2026-10-18T15:14:37.980Z",
          "timeToFirstArtifactMs": 97653,
          "timeToLastArtifactWriteMs": 328582,
          "timeToFirstTokenMs": 24830,
          "firstTurnTtftMs": 23273,
          "timeToFirstToolCallMs": 67498
        },
        "toolCalls": 12,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.53,
        "prefillTokensPerSec": 412.03,
        "decodeTokens": 5084,
        "peakRssMb": 40980,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T15:14:58.372Z",
        "trialId": "interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas",
        "cellId": "r1-qwen3.8-27b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:14:46.925Z",
      "finishedAt": "2026-10-18T15:20:32.874Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 345949,
        "timing": {
          "startedAt": "2026-10-18T15:14:46.925Z",
          "finishedAt": "2026-10-18T15:20:32.874Z",
          "timeToFirstArtifactMs": 338817,
          "timeToLastArtifactWriteMs": 338817,
          "timeToFirstTokenMs": 31219,
          "firstTurnTtftMs": 24631,
          "timeToFirstToolCallMs": 33045
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.08,
        "prefillTokensPerSec": 395.85,
        "decodeTokens": 6114,
        "peakRssMb": 45035,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T15:29:44.567Z",
        "trialId": "craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214",
        "cellId": "r1-qwen3.8-27b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 4292,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 8
        },
        "meanScore": 8,
        "justification": "Grounded summary, severity-ranked findings, concrete fixes and a consistent request-changes verdict follow the review workflow. It catches both SQL injections, file-handle cleanup, pagination, masked errors and debug logging while explicitly respecting all three permitted patterns. Amount validation predates the diff and should be distinguished from regressions; the settings line citation is off.",
        "artifactSha256": "024777b3a87f9f8d6f47e63251b644e75d626e8de0675d1c5af011a60193eec2",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:20:42.276Z",
      "finishedAt": "2026-10-18T15:37:54.575Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1032299,
        "timing": {
          "startedAt": "2026-10-18T15:20:42.276Z",
          "finishedAt": "2026-10-18T15:37:54.575Z",
          "timeToFirstArtifactMs": 244125,
          "timeToLastArtifactWriteMs": 931884,
          "timeToFirstTokenMs": 32808,
          "firstTurnTtftMs": 26297,
          "timeToFirstToolCallMs": 49676
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 19.49,
        "prefillTokensPerSec": 390.2,
        "decodeTokens": 16311,
        "peakRssMb": 43946,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T15:44:55.052Z",
        "trialId": "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5",
        "cellId": "r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 9612,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 6
        },
        "meanScore": 6.5,
        "justification": "Finds the missing history handler, numeric retryCount mismatch and undocumented single-widget GET, while respecting imported session routes and optional nullable email. Checklist, severity and fixes are clear. Several YAML snippets are malformed; Number(undefined) ?? 20 yields NaN, not a default. The claim that no schema has required fields contradicts status/owner, and Revision has no handler.",
        "artifactSha256": "3964da9543a2368557b6327a46c4d79b97b24c4605f7774cff5d5899ac24e032",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:38:06.058Z",
      "finishedAt": "2026-10-18T15:45:23.692Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "craftbook-deep-security-review stale no-write repair loop: security-review/REPORT.md had unchanged content (7487 bytes) for 61s — a re-emission of identical bytes counts as unchanged here after repair feedback and newer chat activity; still failing 9/11: security-review/REPORT.md is missing required content /Verified Safe|Not Statically Verifiable/ — nothing in its 7487 bytes matches. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.8,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 81.8% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 437634,
        "timing": {
          "startedAt": "2026-10-18T15:38:06.058Z",
          "finishedAt": "2026-10-18T15:45:23.692Z",
          "timeToFirstArtifactMs": 326565,
          "timeToLastArtifactWriteMs": 401880,
          "timeToFirstTokenMs": 29301,
          "firstTurnTtftMs": 27668,
          "timeToFirstToolCallMs": 34712
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.79,
        "prefillTokensPerSec": 385.45,
        "decodeTokens": 7894,
        "peakRssMb": 46951,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T16:00:19.764Z",
        "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1",
        "cellId": "r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 7523,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 7,
          "systemic analysis": 6,
          "remediation": 6
        },
        "meanScore": 6.3,
        "justification": "Traces both injections and missing admin authorization, groups root causes, and preserves the three safe exceptions. Report/JSON agree on fixes and severity. It wrongly says other registrars receive requireAdmin rather than requireUser, counts only two of three violated rules, and offers validated request data inside a shell command despite the policy forbidding that.",
        "artifactSha256": "6c291512919058aba9ee5466c8ed3e8e82bc35b2ddd59e1d5dc9bdf6625ac9b4",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 1
      },
      "startedAt": "2026-10-18T15:45:32.687Z",
      "finishedAt": "2026-10-18T17:03:30.705Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4678018,
        "timing": {
          "startedAt": "2026-10-18T15:45:32.687Z",
          "finishedAt": "2026-10-18T17:03:30.705Z",
          "timeToFirstArtifactMs": 2188682,
          "timeToLastArtifactWriteMs": 4668922,
          "timeToFirstTokenMs": 42382,
          "firstTurnTtftMs": 40673,
          "timeToFirstToolCallMs": 45459
        },
        "toolCalls": 48,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 19.82,
        "prefillTokensPerSec": 241.24,
        "decodeTokens": 61380,
        "peakRssMb": 53218,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T17:14:07.113Z",
        "trialId": "large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6",
        "cellId": "r1-qwen3.8-27b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 1
      },
      "startedAt": "2026-10-18T17:03:43.990Z",
      "finishedAt": "2026-10-18T17:19:08.861Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 924871,
        "timing": {
          "startedAt": "2026-10-18T17:03:43.990Z",
          "finishedAt": "2026-10-18T17:19:08.861Z",
          "timeToFirstArtifactMs": 415746,
          "timeToLastArtifactWriteMs": 865930,
          "timeToFirstTokenMs": 32559,
          "firstTurnTtftMs": 25990,
          "timeToFirstToolCallMs": 44964
        },
        "toolCalls": 52,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.05,
        "prefillTokensPerSec": 388.46,
        "decodeTokens": 16224,
        "peakRssMb": 48215,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T17:30:05.109Z",
        "trialId": "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124",
        "cellId": "r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 4083,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 9,
          "Invariant discipline": 9,
          "Reconciled verification": 9
        },
        "meanScore": 9,
        "justification": "Real pre-edit searches found all 7 rename sites and justified 4 exclusions. Captured edits change only the requested identifier and log string; history and behavior remain intact. The 11-to-4 residual ledger reconciles exactly, including re-export and string sites, and the quoted 3/3 test receipt matches the actual tool result.",
        "artifactSha256": "2e9747dbe0928c79e5321bbeb1b68044da3c44d70cb1f2772b92c48bcfa7de56",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 1
      },
      "startedAt": "2026-10-18T17:19:17.638Z",
      "finishedAt": "2026-10-18T17:59:45.987Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2428349,
        "timing": {
          "startedAt": "2026-10-18T17:19:17.638Z",
          "finishedAt": "2026-10-18T17:59:45.987Z",
          "timeToFirstArtifactMs": 391886,
          "timeToLastArtifactWriteMs": 2423777,
          "timeToFirstTokenMs": 25093,
          "firstTurnTtftMs": 18550,
          "timeToFirstToolCallMs": 46126
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.09,
        "prefillTokensPerSec": 392.49,
        "decodeTokens": 46365,
        "peakRssMb": 49801,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T18:15:23.020Z",
        "trialId": "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw",
        "cellId": "r1-qwen3.8-27b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 2465,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 9,
          "Reversible-stage execution": 9,
          "Genuine deduplication": 9
        },
        "meanScore": 9,
        "justification": "Captured runs establish a green 2-test baseline, then 10 characterization tests before restructuring. Each of four reversible stages has a real 12/12 green receipt. Both callers use one unchanged normalization body; all three public exports survive. Source and search receipts substantiate the measured 2-to-1 deduplication.",
        "artifactSha256": "0c07698bdeb300c5589b4001980765db8acd0d2f58134fa86ac58421df48629f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 1
      },
      "startedAt": "2026-10-18T17:59:57.473Z",
      "finishedAt": "2026-10-18T18:09:50.150Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 592677,
        "timing": {
          "startedAt": "2026-10-18T17:59:57.473Z",
          "finishedAt": "2026-10-18T18:09:50.150Z",
          "timeToFirstArtifactMs": 51334,
          "timeToLastArtifactWriteMs": 522589,
          "timeToFirstTokenMs": 34971,
          "firstTurnTtftMs": 28379,
          "timeToFirstToolCallMs": 39936
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.71,
        "prefillTokensPerSec": 403.41,
        "decodeTokens": 9593,
        "peakRssMb": 49002,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T18:15:23.023Z",
        "trialId": "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x",
        "cellId": "r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 2567,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 9,
          "Honest verification": 8
        },
        "meanScore": 8.7,
        "justification": "The real regression test failed with 80 versus 68 before the one-character > to >= fix; a later run passes all four unchanged tests. The patch is general and minimal. Review checks actual source and receipts. Adjacent quantities 9 and 11 are reasoned about rather than directly tested, and fix-notes retains its earlier pending-run caveat.",
        "artifactSha256": "6f0e5b74ee3db6035251964f271b8ffb432f86f01a1d873df4122ece62fc0f40",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 1
      },
      "startedAt": "2026-10-18T18:09:59.804Z",
      "finishedAt": "2026-10-18T18:11:05.900Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 66096,
        "timing": {
          "startedAt": "2026-10-18T18:09:59.804Z",
          "finishedAt": "2026-10-18T18:11:05.900Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 38908,
          "firstTurnTtftMs": 37337,
          "timeToFirstToolCallMs": 43927
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.25,
        "prefillTokensPerSec": 390.9,
        "decodeTokens": 352,
        "peakRssMb": 37549,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T18:14:11.477Z",
        "trialId": "craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 1
      },
      "startedAt": "2026-10-18T18:11:14.804Z",
      "finishedAt": "2026-10-18T18:38:55.462Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1660658,
        "timing": {
          "startedAt": "2026-10-18T18:11:14.804Z",
          "finishedAt": "2026-10-18T18:38:55.462Z",
          "timeToFirstArtifactMs": 1115019,
          "timeToLastArtifactWriteMs": 1329615,
          "timeToFirstTokenMs": 51149,
          "firstTurnTtftMs": 49508,
          "timeToFirstToolCallMs": 58215
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.79,
        "prefillTokensPerSec": 370.31,
        "decodeTokens": 19644,
        "peakRssMb": 46971,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T18:44:05.985Z",
        "trialId": "craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 1
      },
      "startedAt": "2026-10-18T18:39:04.688Z",
      "finishedAt": "2026-10-18T18:49:45.789Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 641101,
        "timing": {
          "startedAt": "2026-10-18T18:39:04.688Z",
          "finishedAt": "2026-10-18T18:49:45.789Z",
          "timeToFirstArtifactMs": 96200,
          "timeToLastArtifactWriteMs": 631845,
          "timeToFirstTokenMs": 32385,
          "firstTurnTtftMs": 25883,
          "timeToFirstToolCallMs": 34452
        },
        "toolCalls": 54,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.53,
        "prefillTokensPerSec": 383.2,
        "decodeTokens": 9651,
        "peakRssMb": 52445,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T19:00:23.250Z",
        "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 1026,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 9,
          "numbering continuity": 9,
          "professionalism": 6,
          "completeness": 4,
          "print-polish": 8
        },
        "meanScore": 7.2,
        "justification": "All three amounts total 5415 and match uninvoiced ledger entries; 042-044 and terms-based due dates are correct. Idle client is explicitly skipped. Clear serif invoices have A4 margins and print rules, but two omit the issue date and none names a currency. Books remain unchanged: lastInvoiceNumber is 041 and all three entries still uninvoiced.",
        "artifactSha256": "9346d3474f5556b3c067df3571d090167379e9e8b039ff6c89b3e5fe74cc7c14",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 1
      },
      "startedAt": "2026-10-18T18:49:55.293Z",
      "finishedAt": "2026-10-18T19:46:11.311Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-linear:6:targetnone:fr1bjnkm5:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 12 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3376018,
        "timing": {
          "startedAt": "2026-10-18T18:49:55.293Z",
          "finishedAt": "2026-10-18T19:46:11.311Z",
          "timeToFirstArtifactMs": 698736,
          "timeToLastArtifactWriteMs": 2289502,
          "timeToFirstTokenMs": 53497,
          "firstTurnTtftMs": 51402,
          "timeToFirstToolCallMs": 73151
        },
        "toolCalls": 56,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 18.46,
        "prefillTokensPerSec": 333.56,
        "decodeTokens": 19732,
        "peakRssMb": 50229,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T19:59:07.939Z",
        "trialId": "craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 1
      },
      "startedAt": "2026-10-18T19:46:23.504Z",
      "finishedAt": "2026-10-18T20:19:17.284Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1973780,
        "timing": {
          "startedAt": "2026-10-18T19:46:23.504Z",
          "finishedAt": "2026-10-18T20:19:17.284Z",
          "timeToFirstArtifactMs": 328075,
          "timeToLastArtifactWriteMs": 1794887,
          "timeToFirstTokenMs": 38980,
          "firstTurnTtftMs": 36876,
          "timeToFirstToolCallMs": 42394
        },
        "toolCalls": 49,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.03,
        "prefillTokensPerSec": 357.32,
        "decodeTokens": 11817,
        "peakRssMb": 50917,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T20:29:10.608Z",
        "trialId": "craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 1
      },
      "startedAt": "2026-10-18T20:19:29.434Z",
      "finishedAt": "2026-10-18T20:41:18.463Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1309029,
        "timing": {
          "startedAt": "2026-10-18T20:19:29.434Z",
          "finishedAt": "2026-10-18T20:41:18.463Z",
          "timeToFirstArtifactMs": 995889,
          "timeToLastArtifactWriteMs": 1307741,
          "timeToFirstTokenMs": 53489,
          "firstTurnTtftMs": 51894,
          "timeToFirstToolCallMs": 156370
        },
        "toolCalls": 24,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 19.31,
        "prefillTokensPerSec": 372.94,
        "decodeTokens": 19415,
        "peakRssMb": 47546,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T20:44:06.196Z",
        "trialId": "craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 1
      },
      "startedAt": "2026-10-18T20:41:27.919Z",
      "finishedAt": "2026-10-18T21:25:13.374Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2625455,
        "timing": {
          "startedAt": "2026-10-18T20:41:27.919Z",
          "finishedAt": "2026-10-18T21:25:13.374Z",
          "timeToFirstArtifactMs": 1762877,
          "timeToLastArtifactWriteMs": 2619616,
          "timeToFirstTokenMs": 57054,
          "firstTurnTtftMs": 54967,
          "timeToFirstToolCallMs": 66871
        },
        "toolCalls": 46,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 19.19,
        "prefillTokensPerSec": 319.33,
        "decodeTokens": 15110,
        "peakRssMb": 48108,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T21:29:11.576Z",
        "trialId": "craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 1
      },
      "startedAt": "2026-10-18T21:25:22.821Z",
      "finishedAt": "2026-10-18T21:46:59.980Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 603s — model busy but not delivering (hard digest stuck at a077d58974c1b070)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_NO_PRODUCED_OUTPUT",
            "summary": "No authored artifact event or scored sniff output was observed."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1297159,
        "timing": {
          "startedAt": "2026-10-18T21:25:22.821Z",
          "finishedAt": "2026-10-18T21:46:59.980Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 55366,
          "firstTurnTtftMs": 53789,
          "timeToFirstToolCallMs": 69932
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 15.3,
        "prefillTokensPerSec": 258.58,
        "decodeTokens": 5744,
        "peakRssMb": 43344,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T21:59:04.373Z",
        "trialId": "craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za",
      "cell": {
        "id": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 1
      },
      "startedAt": "2026-10-18T21:47:11.729Z",
      "finishedAt": "2026-10-18T23:10:26.666Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-fanout:6:targetnone:fr1tjwffw:rp0:rf0:m1\" stuck for 18m — artifact scored but stalled 18m despite 10 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4994937,
        "timing": {
          "startedAt": "2026-10-18T21:47:11.729Z",
          "finishedAt": "2026-10-18T23:10:26.666Z",
          "timeToFirstArtifactMs": 2207575,
          "timeToLastArtifactWriteMs": 4021706,
          "timeToFirstTokenMs": 56095,
          "firstTurnTtftMs": 54006,
          "timeToFirstToolCallMs": 71187
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 15.8,
        "prefillTokensPerSec": 279.31,
        "decodeTokens": 33546,
        "peakRssMb": 49967,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:15:00.305Z",
        "trialId": "craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za",
        "cellId": "r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:10:36.972Z",
      "finishedAt": "2026-10-18T23:13:28.113Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 171141,
        "timing": {
          "startedAt": "2026-10-18T23:10:36.972Z",
          "finishedAt": "2026-10-18T23:13:28.113Z",
          "timeToFirstArtifactMs": 163527,
          "timeToLastArtifactWriteMs": 163531,
          "timeToFirstTokenMs": 26977,
          "firstTurnTtftMs": 25044,
          "timeToFirstToolCallMs": 38860
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.41,
        "prefillTokensPerSec": 370.59,
        "decodeTokens": 2871,
        "peakRssMb": 41741,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:17:57.936Z",
        "trialId": "tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj",
        "cellId": "r2-qwen3.8-27b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 3553,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "Clean dark board with distinct player colors and winning-line highlights. Browser checks confirm row, column and diagonal wins, draw, move locking, reset and a 320px fit. State handling is readable. Empty cells have no accessible names and turn/win status is not announced; presentation remains a basic fixed-size board.",
        "artifactSha256": "14da66f038c031a8d0193c450035f3382126c6e94d2bfa2c321076235f51f7ae",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:14:02.858Z",
      "finishedAt": "2026-10-18T23:17:24.958Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 202100,
        "timing": {
          "startedAt": "2026-10-18T23:14:02.858Z",
          "finishedAt": "2026-10-18T23:17:24.958Z",
          "timeToFirstArtifactMs": 196283,
          "timeToLastArtifactWriteMs": 196291,
          "timeToFirstTokenMs": 24731,
          "firstTurnTtftMs": 23182,
          "timeToFirstToolCallMs": 46157
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.64,
        "prefillTokensPerSec": 413.07,
        "decodeTokens": 4616,
        "peakRssMb": 47884,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:21:06.327Z",
        "trialId": "petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4",
        "cellId": "r2-qwen3.8-27b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/workspace/pet-shop-website/index.html",
        "artifactBytes": 9492,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 7,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.3,
        "justification": "Warm, readable product/service cards and responsive layout fit 360px. The real generated PNG is wired correctly, but its abstract imagery and garbled lettering make a weak shop logo. Labels, live form status and validation work; a filled form falsely promises a reply while only resetting locally. Product imagery remains emoji.",
        "artifactSha256": "6837358048a683cfcb81f02310625b52562dac22cc15d8b01c15f128d666602b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:17:58.673Z",
      "finishedAt": "2026-10-18T23:21:24.664Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 205991,
        "timing": {
          "startedAt": "2026-10-18T23:17:58.673Z",
          "finishedAt": "2026-10-18T23:21:24.664Z",
          "timeToFirstArtifactMs": 198129,
          "timeToLastArtifactWriteMs": 198133,
          "timeToFirstTokenMs": 25360,
          "firstTurnTtftMs": 23820,
          "timeToFirstToolCallMs": 70601
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 30.02,
        "prefillTokensPerSec": 440.96,
        "decodeTokens": 4808,
        "peakRssMb": 39229,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:25:10.144Z",
        "trialId": "tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c",
        "cellId": "r2-qwen3.8-27b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 9313,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 6
        },
        "meanScore": 6.8,
        "justification": "Readable grid arena, colored tanks and score/lives/wave HUD. Source implements aimed fire from both sides; browser probes confirm movement, hit scoring, wave progression, damage and restart. Delta-time movement and normalized diagonals are sound. The fixed 800px canvas lacks responsive scaling, blur can leave keys held, and combat has little hit feedback.",
        "artifactSha256": "eef18b5642f3504173b3d8490943dbff455a93dbf1e0915bdfabb0419379e43e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:22:00.558Z",
      "finishedAt": "2026-10-18T23:25:59.639Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 239081,
        "timing": {
          "startedAt": "2026-10-18T23:22:00.558Z",
          "finishedAt": "2026-10-18T23:25:59.639Z",
          "timeToFirstArtifactMs": 24109,
          "timeToLastArtifactWriteMs": 234558,
          "timeToFirstTokenMs": 13866,
          "firstTurnTtftMs": 12277,
          "timeToFirstToolCallMs": 24111
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 29.14,
        "prefillTokensPerSec": 442.55,
        "decodeTokens": 3289,
        "peakRssMb": 47785,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:26:52.515Z",
        "trialId": "schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0",
        "cellId": "r2-qwen3.8-27b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:26:09.472Z",
      "finishedAt": "2026-10-18T23:27:02.553Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 53081,
        "timing": {
          "startedAt": "2026-10-18T23:26:09.472Z",
          "finishedAt": "2026-10-18T23:27:02.553Z",
          "timeToFirstArtifactMs": 50611,
          "timeToLastArtifactWriteMs": 50612,
          "timeToFirstTokenMs": 18156,
          "firstTurnTtftMs": 16542,
          "timeToFirstToolCallMs": 19590
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 30.38,
        "prefillTokensPerSec": 389.46,
        "decodeTokens": 614,
        "peakRssMb": 37718,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:28:11.146Z",
        "trialId": "failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a",
        "cellId": "r2-qwen3.8-27b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:27:11.492Z",
      "finishedAt": "2026-10-18T23:27:54.144Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 42652,
        "timing": {
          "startedAt": "2026-10-18T23:27:11.492Z",
          "finishedAt": "2026-10-18T23:27:54.144Z",
          "timeToFirstArtifactMs": 41165,
          "timeToLastArtifactWriteMs": 41165,
          "timeToFirstTokenMs": 18272,
          "firstTurnTtftMs": 16678,
          "timeToFirstToolCallMs": 20011
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.25,
        "prefillTokensPerSec": 396.67,
        "decodeTokens": 406,
        "peakRssMb": 37110,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:28:11.147Z",
        "trialId": "symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66",
        "cellId": "r2-qwen3.8-27b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:28:03.443Z",
      "finishedAt": "2026-10-18T23:29:41.410Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 97967,
        "timing": {
          "startedAt": "2026-10-18T23:28:03.443Z",
          "finishedAt": "2026-10-18T23:29:41.410Z",
          "timeToFirstArtifactMs": 74946,
          "timeToLastArtifactWriteMs": 91988,
          "timeToFirstTokenMs": 18338,
          "firstTurnTtftMs": 16741,
          "timeToFirstToolCallMs": 21325
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.83,
        "prefillTokensPerSec": 390.15,
        "decodeTokens": 1828,
        "peakRssMb": 38306,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:30:56.613Z",
        "trialId": "data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo",
        "cellId": "r2-qwen3.8-27b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:29:50.613Z",
      "finishedAt": "2026-10-18T23:40:30.813Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 640200,
        "timing": {
          "startedAt": "2026-10-18T23:29:50.613Z",
          "finishedAt": "2026-10-18T23:40:30.813Z",
          "timeToFirstArtifactMs": 518968,
          "timeToLastArtifactWriteMs": 637746,
          "timeToFirstTokenMs": 35131,
          "firstTurnTtftMs": 33538,
          "timeToFirstToolCallMs": 47433
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.27,
        "prefillTokensPerSec": 371.69,
        "decodeTokens": 11871,
        "peakRssMb": 46232,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:45:51.527Z",
        "trialId": "incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8",
        "cellId": "r2-qwen3.8-27b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:40:40.174Z",
      "finishedAt": "2026-10-18T23:42:37.883Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 117709,
        "timing": {
          "startedAt": "2026-10-18T23:40:40.174Z",
          "finishedAt": "2026-10-18T23:42:37.883Z",
          "timeToFirstArtifactMs": 48862,
          "timeToLastArtifactWriteMs": 115158,
          "timeToFirstTokenMs": 36799,
          "firstTurnTtftMs": 35198,
          "timeToFirstToolCallMs": 39333
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.57,
        "prefillTokensPerSec": 392.27,
        "decodeTokens": 1172,
        "peakRssMb": 41395,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:45:51.528Z",
        "trialId": "ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t",
        "cellId": "r2-qwen3.8-27b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:42:47.488Z",
      "finishedAt": "2026-10-18T23:45:10.331Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 142843,
        "timing": {
          "startedAt": "2026-10-18T23:42:47.488Z",
          "finishedAt": "2026-10-18T23:45:10.331Z",
          "timeToFirstArtifactMs": 140639,
          "timeToLastArtifactWriteMs": 140639,
          "timeToFirstTokenMs": 18988,
          "firstTurnTtftMs": 17402,
          "timeToFirstToolCallMs": 140641
        },
        "toolCalls": 1,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.66,
        "prefillTokensPerSec": 408.59,
        "decodeTokens": 2148,
        "peakRssMb": 36814,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-18T23:45:51.528Z",
        "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m",
        "cellId": "r2-qwen3.8-27b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:45:19.891Z",
      "finishedAt": "2026-10-18T23:50:34.032Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 314141,
        "timing": {
          "startedAt": "2026-10-18T23:45:19.891Z",
          "finishedAt": "2026-10-18T23:50:34.032Z",
          "timeToFirstArtifactMs": 274978,
          "timeToLastArtifactWriteMs": 307933,
          "timeToFirstTokenMs": 35954,
          "firstTurnTtftMs": 34356,
          "timeToFirstToolCallMs": 43371
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.56,
        "prefillTokensPerSec": 379.55,
        "decodeTokens": 5378,
        "peakRssMb": 45869,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:00:44.906Z",
        "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw",
        "cellId": "r2-qwen3.8-27b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:50:42.989Z",
      "finishedAt": "2026-10-18T23:52:10.572Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 87583,
        "timing": {
          "startedAt": "2026-10-18T23:50:42.989Z",
          "finishedAt": "2026-10-18T23:52:10.572Z",
          "timeToFirstArtifactMs": 24331,
          "timeToLastArtifactWriteMs": 85283,
          "timeToFirstTokenMs": 14092,
          "firstTurnTtftMs": 12506,
          "timeToFirstToolCallMs": 24334
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.65,
        "prefillTokensPerSec": 412.4,
        "decodeTokens": 898,
        "peakRssMb": 40702,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:00:44.907Z",
        "trialId": "constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut",
        "cellId": "r2-qwen3.8-27b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 2
      },
      "startedAt": "2026-10-18T23:52:20.247Z",
      "finishedAt": "2026-10-19T00:02:46.115Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 625868,
        "timing": {
          "startedAt": "2026-10-18T23:52:20.247Z",
          "finishedAt": "2026-10-19T00:02:46.115Z",
          "timeToFirstArtifactMs": 106303,
          "timeToLastArtifactWriteMs": 622982,
          "timeToFirstTokenMs": 19022,
          "firstTurnTtftMs": 17437,
          "timeToFirstToolCallMs": 20825
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.78,
        "prefillTokensPerSec": 360.05,
        "decodeTokens": 14052,
        "peakRssMb": 40809,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:03:49.889Z",
        "trialId": "records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68",
        "cellId": "r2-qwen3.8-27b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:02:55.715Z",
      "finishedAt": "2026-10-19T00:05:53.794Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 178079,
        "timing": {
          "startedAt": "2026-10-19T00:02:55.715Z",
          "finishedAt": "2026-10-19T00:05:53.794Z",
          "timeToFirstArtifactMs": 173695,
          "timeToLastArtifactWriteMs": 173695,
          "timeToFirstTokenMs": 19612,
          "firstTurnTtftMs": 18020,
          "timeToFirstToolCallMs": 88045
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 16.95,
        "prefillTokensPerSec": 362.41,
        "decodeTokens": 2614,
        "peakRssMb": 37107,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:17:05.539Z",
        "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4",
        "cellId": "r2-qwen3.8-27b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:06:03.261Z",
      "finishedAt": "2026-10-19T00:08:41.599Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 158338,
        "timing": {
          "startedAt": "2026-10-19T00:06:03.261Z",
          "finishedAt": "2026-10-19T00:08:41.599Z",
          "timeToFirstArtifactMs": 53493,
          "timeToLastArtifactWriteMs": 157234,
          "timeToFirstTokenMs": 26548,
          "firstTurnTtftMs": 24966,
          "timeToFirstToolCallMs": 29444
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 34.34,
        "prefillTokensPerSec": 421.22,
        "decodeTokens": 1085,
        "peakRssMb": 38786,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:21:53.446Z",
        "trialId": "meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz",
        "cellId": "r2-qwen3.8-27b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1590,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 8,
          "grounding": 9,
          "actionability": 9,
          "brevity": 8
        },
        "meanScore": 8.5,
        "justification": "Clearly preserves Sep 14/EU/overnight decisions, annual-plan and rollback unknowns, all four owners/dates, and A2 dependency. A4 correctly schedules the review by Aug 30. Legal risk is explicit; stale proposals are rejected. Compact table is executable, though scheduling also appears under decisions and risks repeat open questions.",
        "artifactSha256": "e05dd21ec2f26ed4970acec911de00af641d64bda71fb4e134a76ee4ca7dd501",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:08:51.052Z",
      "finishedAt": "2026-10-19T00:26:07.802Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1036750,
        "timing": {
          "startedAt": "2026-10-19T00:08:51.052Z",
          "finishedAt": "2026-10-19T00:26:07.802Z",
          "timeToFirstArtifactMs": 100307,
          "timeToLastArtifactWriteMs": 505149,
          "timeToFirstTokenMs": 25283,
          "firstTurnTtftMs": 18749,
          "timeToFirstToolCallMs": 35896
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.87,
        "prefillTokensPerSec": 423.04,
        "decodeTokens": 20979,
        "peakRssMb": 42781,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:31:12.079Z",
        "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh",
        "cellId": "r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 3939,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 6
        },
        "meanScore": 6.5,
        "justification": "Scope→build→evaluate→finish and a clear 13-driver contract; live totals and flat sanity work. Reviewer arithmetic is wrong ($48,434 vs actual $49,393). Indexing across 13-cell rows shifts monthly columns: M1 revenue shows $4,609, should $4,760. Scope also assumes a monthly source period absent from records.",
        "artifactSha256": "f0ce8ab1f0a162d80f72d448cd775f4c903a9dd614b7956b86cbb632146ed098",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:26:17.350Z",
      "finishedAt": "2026-10-19T00:32:46.619Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 389269,
        "timing": {
          "startedAt": "2026-10-19T00:26:17.350Z",
          "finishedAt": "2026-10-19T00:32:46.619Z",
          "timeToFirstArtifactMs": 186643,
          "timeToLastArtifactWriteMs": 383210,
          "timeToFirstTokenMs": 36849,
          "firstTurnTtftMs": 35253,
          "timeToFirstToolCallMs": 43490
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.73,
        "prefillTokensPerSec": 377.98,
        "decodeTokens": 5266,
        "peakRssMb": 46130,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:45:06.178Z",
        "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix",
        "cellId": "r2-qwen3.8-27b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:32:55.783Z",
      "finishedAt": "2026-10-19T00:34:38.739Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 102956,
        "timing": {
          "startedAt": "2026-10-19T00:32:55.783Z",
          "finishedAt": "2026-10-19T00:34:38.739Z",
          "timeToFirstArtifactMs": 98188,
          "timeToLastArtifactWriteMs": 98188,
          "timeToFirstTokenMs": 23913,
          "firstTurnTtftMs": 22229,
          "timeToFirstToolCallMs": 24631
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.53,
        "prefillTokensPerSec": 368.61,
        "decodeTokens": 1660,
        "peakRssMb": 43299,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T00:45:54.429Z",
        "trialId": "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89",
        "cellId": "r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1783,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 8,
          "candor": 9,
          "usefulness": 8
        },
        "meanScore": 8.3,
        "justification": "Theme values, 11→9 pages, spacing note and preserved content match captured tool responses. Callout Emphasis is plainly unresolved, with actionable brand-style and interim-restyling options. Minor count error: five styles already transferred; adding the missing Callout would make six, not “all five.”",
        "artifactSha256": "151d95e67a3242c088528c1bf7716100a684875f01ef16e4c598d6c416c5a96a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:34:48.043Z",
      "finishedAt": "2026-10-19T00:48:59.997Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 851954,
        "timing": {
          "startedAt": "2026-10-19T00:34:48.043Z",
          "finishedAt": "2026-10-19T00:48:59.997Z",
          "timeToFirstArtifactMs": 248234,
          "timeToLastArtifactWriteMs": 845495,
          "timeToFirstTokenMs": 39126,
          "firstTurnTtftMs": 32495,
          "timeToFirstToolCallMs": 42042
        },
        "toolCalls": 32,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.87,
        "prefillTokensPerSec": 378.35,
        "decodeTokens": 14197,
        "peakRssMb": 48342,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T01:00:30.888Z",
        "trialId": "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem",
        "cellId": "r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 2864,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 7,
          "convertibility": 7
        },
        "meanScore": 8,
        "justification": "Read and cited the exact brief before outlining; all seven outline titles appear once in order. Metrics and three actions are preserved; real convert/preview/save/copy receipts exist. “Controlled, measurable test” and “approach is working” exceed the sparse evidence. Repeated no-invention caveats clutter otherwise convertible slide blocks.",
        "artifactSha256": "edb4b6377a67cee3263f523a804923ae77a0d99f0b66633e6ef71c03dc1d6078",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 2
      },
      "startedAt": "2026-10-19T00:49:09.635Z",
      "finishedAt": "2026-10-19T01:00:25.107Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 675472,
        "timing": {
          "startedAt": "2026-10-19T00:49:09.635Z",
          "finishedAt": "2026-10-19T01:00:25.107Z",
          "timeToFirstArtifactMs": 590424,
          "timeToLastArtifactWriteMs": 672223,
          "timeToFirstTokenMs": 15795,
          "firstTurnTtftMs": 14145,
          "timeToFirstToolCallMs": 90367
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 19.96,
        "prefillTokensPerSec": 366.57,
        "decodeTokens": 11535,
        "peakRssMb": 42178,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T01:01:08.028Z",
        "trialId": "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip",
        "cellId": "r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 6585,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 5,
          "grounding": 6,
          "citationQuality": 8,
          "calibration": 7
        },
        "meanScore": 6.5,
        "justification": "Correct chronology, near-claim source IDs and explicit debate over “first programmer”; final brief excludes the band citation. Repeats short cards across summary, timeline and findings instead of building much argument. “Never executed on working hardware” and “most often cited” exceed the signed evidence; non-completion alone cannot establish those claims.",
        "artifactSha256": "a459ec85e0abee96df48f1ac6fe3d699ba5171cf8588794e77d3a5454cbfd2be",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 2
      },
      "startedAt": "2026-10-19T01:00:34.340Z",
      "finishedAt": "2026-10-19T02:30:41.493Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "hit hard ceiling (2700000ms + 45m progress extensions) — last hard progress 176s before the end; the deliverable never closed",
        "failureMode": "timeout",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 0,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 5407153,
        "timing": {
          "startedAt": "2026-10-19T01:00:34.340Z",
          "finishedAt": "2026-10-19T02:30:41.493Z",
          "timeToFirstArtifactMs": 458483,
          "timeToLastArtifactWriteMs": 3553144,
          "timeToFirstTokenMs": 49035,
          "firstTurnTtftMs": 42480,
          "timeToFirstToolCallMs": 53005
        },
        "toolCalls": 142,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.04,
        "prefillTokensPerSec": 393.54,
        "decodeTokens": 85147,
        "peakRssMb": 50241,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T02:45:42.719Z",
        "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v",
        "cellId": "r2-qwen3.8-27b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 5154,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 1,
          "prioritization": 1,
          "conflict-handling": 0,
          "realism": 1
        },
        "meanScore": 0.8,
        "justification": "Calendar calls are empty. Replaces eight July events with eleven October meetings, omits Q3 budget/all-hands, and misses the actual budget/Priya overlap. Focus gaps and heavy-day claims rest on invented times. SOW is after the demo; even its suggested 10:30 alternative is after 09:30. Admitting failed retrieval does not ground this plan.",
        "artifactSha256": "8981e4cc8f32732b167192cb509c8bec809e4077d914bab55023141f49a90d0f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 2
      },
      "startedAt": "2026-10-19T02:30:54.417Z",
      "finishedAt": "2026-10-19T03:15:19.661Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2665244,
        "timing": {
          "startedAt": "2026-10-19T02:30:54.417Z",
          "finishedAt": "2026-10-19T03:15:19.661Z",
          "timeToFirstArtifactMs": 350820,
          "timeToLastArtifactWriteMs": 2662606,
          "timeToFirstTokenMs": 34314,
          "firstTurnTtftMs": 27795,
          "timeToFirstToolCallMs": 52019
        },
        "toolCalls": 52,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.44,
        "prefillTokensPerSec": 418.35,
        "decodeTokens": 50665,
        "peakRssMb": 48605,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:30:10.441Z",
        "trialId": "craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri",
        "cellId": "r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2835,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:15:29.762Z",
      "finishedAt": "2026-10-19T03:23:56.309Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 506547,
        "timing": {
          "startedAt": "2026-10-19T03:15:29.762Z",
          "finishedAt": "2026-10-19T03:23:56.309Z",
          "timeToFirstArtifactMs": 90536,
          "timeToLastArtifactWriteMs": 380489,
          "timeToFirstTokenMs": 33780,
          "firstTurnTtftMs": 27256,
          "timeToFirstToolCallMs": 48573
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.89,
        "prefillTokensPerSec": 369.99,
        "decodeTokens": 7824,
        "peakRssMb": 47565,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:30:44.324Z",
        "trialId": "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1",
        "cellId": "r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 6202,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 7,
          "grounding": 4,
          "structure": 8,
          "actionability": 6
        },
        "meanScore": 6.3,
        "justification": "All six entries retain seeded numbers and uniform APA-style Summary/Evaluation/Relevance sections, with missing metadata acknowledged. Adds unsupported facts: Allen had no control group, Chen was cross-sectional, Dunn had adequate power for large effects, and Williams proposed a checklist. Patel’s hybrid mix and claims about unreported outcomes also exceed the cards.",
        "artifactSha256": "ac15fd928dd83f0e7239a9dca924be9afc539ce24ad635af7e93379ab7809780",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:24:05.689Z",
      "finishedAt": "2026-10-19T03:33:22.406Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 556717,
        "timing": {
          "startedAt": "2026-10-19T03:24:05.689Z",
          "finishedAt": "2026-10-19T03:33:22.406Z",
          "timeToFirstArtifactMs": 79977,
          "timeToLastArtifactWriteMs": 338338,
          "timeToFirstTokenMs": 32106,
          "firstTurnTtftMs": 25579,
          "timeToFirstToolCallMs": 35089
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.92,
        "prefillTokensPerSec": 371.85,
        "decodeTokens": 7553,
        "peakRssMb": 49780,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:45:56.215Z",
        "trialId": "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi",
        "cellId": "r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 2571,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 4,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 6.7,
        "justification": "Correct seeded numbers and three actions, with a usable summary/table/sections and plain convertible Markdown. Adds unsupported complaint-volume trends, escalation reduction, “only” unresolved gap and largest-improvement claims. Treats 18→6 hours as three working days→one, and assigns all actions the same status-silence purpose without source support.",
        "artifactSha256": "4db3e0bd520fe81a1c98b98431b32ad1404a507bca79344c8bbfaaaf0c264350",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:33:31.871Z",
      "finishedAt": "2026-10-19T03:35:51.893Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 140022,
        "timing": {
          "startedAt": "2026-10-19T03:33:31.871Z",
          "finishedAt": "2026-10-19T03:35:51.893Z",
          "timeToFirstArtifactMs": 78960,
          "timeToLastArtifactWriteMs": 136460,
          "timeToFirstTokenMs": 19307,
          "firstTurnTtftMs": 17694,
          "timeToFirstToolCallMs": 23722
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.08,
        "prefillTokensPerSec": 352.71,
        "decodeTokens": 2929,
        "peakRssMb": 37427,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:45:05.754Z",
        "trialId": "fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol",
        "cellId": "r2-qwen3.8-27b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:36:01.232Z",
      "finishedAt": "2026-10-19T03:37:19.390Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 78158,
        "timing": {
          "startedAt": "2026-10-19T03:36:01.232Z",
          "finishedAt": "2026-10-19T03:37:19.390Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 38341,
          "firstTurnTtftMs": 36769,
          "timeToFirstToolCallMs": 47859
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.47,
        "prefillTokensPerSec": 394.17,
        "decodeTokens": 677,
        "peakRssMb": 37539,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:45:05.754Z",
        "trialId": "dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4",
        "cellId": "r2-qwen3.8-27b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:37:28.341Z",
      "finishedAt": "2026-10-19T03:39:26.079Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 5 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 117738,
        "timing": {
          "startedAt": "2026-10-19T03:37:28.341Z",
          "finishedAt": "2026-10-19T03:39:26.079Z",
          "timeToFirstArtifactMs": 113911,
          "timeToLastArtifactWriteMs": 113926,
          "timeToFirstTokenMs": 25788,
          "firstTurnTtftMs": 24240,
          "timeToFirstToolCallMs": 53108
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.53,
        "prefillTokensPerSec": 427.54,
        "decodeTokens": 1730,
        "peakRssMb": 39251,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:45:05.754Z",
        "trialId": "interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx",
        "cellId": "r2-qwen3.8-27b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:39:35.564Z",
      "finishedAt": "2026-10-19T03:42:57.295Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 201731,
        "timing": {
          "startedAt": "2026-10-19T03:39:35.564Z",
          "finishedAt": "2026-10-19T03:42:57.295Z",
          "timeToFirstArtifactMs": 195586,
          "timeToLastArtifactWriteMs": 195586,
          "timeToFirstTokenMs": 30127,
          "firstTurnTtftMs": 23598,
          "timeToFirstToolCallMs": 31855
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.29,
        "prefillTokensPerSec": 407.97,
        "decodeTokens": 2770,
        "peakRssMb": 42324,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T03:45:56.216Z",
        "trialId": "craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667",
        "cellId": "r2-qwen3.8-27b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3717,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 7
        },
        "meanScore": 7.5,
        "justification": "Reads snapshot and scopes house rules; catches SQL regressions, resource leak, false-success catch, pagination and debug log without flagging allowed patterns. Fixes are specific. Amount validation was already absent before this diff, so a new major finding overstates scope. Capture reaches Done but contains no final verdict note.",
        "artifactSha256": "2a5fc2590ba288758a1b571ee96ee0e600dc32af10567caab776c1830ac6238a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 2
      },
      "startedAt": "2026-10-19T03:43:06.201Z",
      "finishedAt": "2026-10-19T04:14:52.628Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1906427,
        "timing": {
          "startedAt": "2026-10-19T03:43:06.201Z",
          "finishedAt": "2026-10-19T04:14:52.628Z",
          "timeToFirstArtifactMs": 326651,
          "timeToLastArtifactWriteMs": 1706929,
          "timeToFirstTokenMs": 33853,
          "firstTurnTtftMs": 27317,
          "timeToFirstToolCallMs": 47023
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.08,
        "prefillTokensPerSec": 400.24,
        "decodeTokens": 35888,
        "peakRssMb": 47567,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T04:15:34.636Z",
        "trialId": "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe",
        "cellId": "r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 13267,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 6
        },
        "meanScore": 7,
        "justification": "Inventories both route files; catches missing history, handler-only GET and numeric retryCount while accepting optional nullable email. Snippets are actionable. Absence of version history does not establish a NEW API or prove compatibility; the report wrongly guarantees no deployed consumers. Verb naming is inflated to a blocker and several risks become certainties.",
        "artifactSha256": "6e4e04ea1dfba83e5d5d673f382c904765dded1d6bcd403c3903b7e6c4883d4d",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 2
      },
      "startedAt": "2026-10-19T04:15:02.945Z",
      "finishedAt": "2026-10-19T04:20:54.247Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 351302,
        "timing": {
          "startedAt": "2026-10-19T04:15:02.945Z",
          "finishedAt": "2026-10-19T04:20:54.247Z",
          "timeToFirstArtifactMs": 98870,
          "timeToLastArtifactWriteMs": 347095,
          "timeToFirstTokenMs": 26958,
          "firstTurnTtftMs": 25323,
          "timeToFirstToolCallMs": 32009
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.83,
        "prefillTokensPerSec": 358.78,
        "decodeTokens": 4242,
        "peakRssMb": 46267,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T04:30:34.215Z",
        "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn",
        "cellId": "r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 5702,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 7,
          "remediation": 7
        },
        "meanScore": 7.3,
        "justification": "Confirms all three admin source-to-sink/auth paths with matching JSON and correctly exempts profile, health, static exec and allowlisted sorting. Fixes and admin regression tests are concrete. Claims about absent CI gates exceed inspected sources; proposed blanket rejection of ${ in SQL would wrongly flag the explicitly safe ORDER BY pattern.",
        "artifactSha256": "9b3dc68fcac561de78d2df027adabc3ffe98a45016ef148c52d28ffa5f56854b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 2
      },
      "startedAt": "2026-10-19T04:21:03.208Z",
      "finishedAt": "2026-10-19T05:04:47.535Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 723s — model busy but not delivering (hard digest stuck at 5b397031a20518e6)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_UNGRADED_AUTHORED_OUTPUT",
            "summary": "An authored output exists without a sniff verdict; quality is capped at the rubric midpoint."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2624327,
        "timing": {
          "startedAt": "2026-10-19T04:21:03.208Z",
          "finishedAt": "2026-10-19T05:04:47.535Z",
          "timeToFirstArtifactMs": 1882554,
          "timeToLastArtifactWriteMs": 1952170,
          "timeToFirstTokenMs": 44621,
          "firstTurnTtftMs": 42905,
          "timeToFirstToolCallMs": 47711
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.91,
        "prefillTokensPerSec": 206.17,
        "decodeTokens": 21215,
        "peakRssMb": 52906,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T05:15:06.169Z",
        "trialId": "large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862",
        "cellId": "r2-qwen3.8-27b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 2
      },
      "startedAt": "2026-10-19T05:04:57.177Z",
      "finishedAt": "2026-10-19T05:18:40.788Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 823611,
        "timing": {
          "startedAt": "2026-10-19T05:04:57.177Z",
          "finishedAt": "2026-10-19T05:18:40.788Z",
          "timeToFirstArtifactMs": 247117,
          "timeToLastArtifactWriteMs": 763482,
          "timeToFirstTokenMs": 33992,
          "firstTurnTtftMs": 27435,
          "timeToFirstToolCallMs": 35923
        },
        "toolCalls": 58,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.81,
        "prefillTokensPerSec": 378.25,
        "decodeTokens": 13724,
        "peakRssMb": 49614,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T05:30:54.298Z",
        "trialId": "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc",
        "cellId": "r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 3577,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 9,
          "Invariant discipline": 9,
          "Reconciled verification": 9
        },
        "meanScore": 9,
        "justification": "Real pre-edit ripgrep finds 11 lines; seven scoped rename sites include re-export and template log, with four justified exclusions. Recorded replacements change only those sites. Post-edit grep returns exactly four excluded lines; later full reads confirm them. Baseline and post-edit three-test receipts are real and green; 7+4=11 reconciles.",
        "artifactSha256": "718a01d17ad3ff075fc4789574a5ed6d725fcf7a4c1815b1ec77bf3cfa03f989",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 2
      },
      "startedAt": "2026-10-19T05:18:50.472Z",
      "finishedAt": "2026-10-19T05:55:05.372Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2174900,
        "timing": {
          "startedAt": "2026-10-19T05:18:50.472Z",
          "finishedAt": "2026-10-19T05:55:05.372Z",
          "timeToFirstArtifactMs": 488508,
          "timeToLastArtifactWriteMs": 2051895,
          "timeToFirstTokenMs": 25703,
          "firstTurnTtftMs": 19162,
          "timeToFirstToolCallMs": 54680
        },
        "toolCalls": 74,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.49,
        "prefillTokensPerSec": 417.48,
        "decodeTokens": 37900,
        "peakRssMb": 50314,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T06:00:47.554Z",
        "trialId": "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek",
        "cellId": "r2-qwen3.8-27b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 3003,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 9,
          "Reversible-stage execution": 9,
          "Genuine deduplication": 9
        },
        "meanScore": 9,
        "justification": "Real 2/2 baseline precedes nine characterization tests and an 11/11 run before source edits. Three actual stages add helper, delegate formatName, then replace the private report helper; each has an intervening green 11-test receipt. One verbatim normalizer remains, both callers use it, public exports survive. Line-count estimate deviations are explicitly recorded.",
        "artifactSha256": "3f6fa5bd2dfdb76ca1aed6629c6cd1d81b04c17faf3cf6a1f450f1bc7cf38780",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 2
      },
      "startedAt": "2026-10-19T05:55:16.430Z",
      "finishedAt": "2026-10-19T06:04:43.141Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 566711,
        "timing": {
          "startedAt": "2026-10-19T05:55:16.430Z",
          "finishedAt": "2026-10-19T06:04:43.141Z",
          "timeToFirstArtifactMs": 39140,
          "timeToLastArtifactWriteMs": 492401,
          "timeToFirstTokenMs": 28710,
          "firstTurnTtftMs": 22141,
          "timeToFirstToolCallMs": 31024
        },
        "toolCalls": 40,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.08,
        "prefillTokensPerSec": 405.53,
        "decodeTokens": 8100,
        "peakRssMb": 48731,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T06:17:42.391Z",
        "trialId": "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv",
        "cellId": "r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 2402,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 9,
          "Honest verification": 8
        },
        "meanScore": 8.7,
        "justification": "A real boundary test failed with 80 !== 68 before the single > to >= edit; the original test then passed with all four tests green. The fix preserves validation and rounding, and the review checks the actual files and receipts. The notes candidly identify missing adjacent-boundary coverage; validation’s claim that nothing remains unverified is broader than the tests establish.",
        "artifactSha256": "cfa9971fe0c3f90a1466a19b587286aaf2bf998cd93ca63f650898417c9ce6a1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 2
      },
      "startedAt": "2026-10-19T06:04:52.686Z",
      "finishedAt": "2026-10-19T06:06:03.713Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 71027,
        "timing": {
          "startedAt": "2026-10-19T06:04:52.686Z",
          "finishedAt": "2026-10-19T06:06:03.713Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 39930,
          "firstTurnTtftMs": 38354,
          "timeToFirstToolCallMs": 45226
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.28,
        "prefillTokensPerSec": 378.78,
        "decodeTokens": 397,
        "peakRssMb": 37512,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T06:15:11.750Z",
        "trialId": "craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 2
      },
      "startedAt": "2026-10-19T06:06:12.771Z",
      "finishedAt": "2026-10-19T06:24:04.542Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1071771,
        "timing": {
          "startedAt": "2026-10-19T06:06:12.771Z",
          "finishedAt": "2026-10-19T06:24:04.542Z",
          "timeToFirstArtifactMs": 632898,
          "timeToLastArtifactWriteMs": 972716,
          "timeToFirstTokenMs": 52104,
          "firstTurnTtftMs": 50353,
          "timeToFirstToolCallMs": 59753
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 18.86,
        "prefillTokensPerSec": 347.54,
        "decodeTokens": 12310,
        "peakRssMb": 46124,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T06:30:10.953Z",
        "trialId": "craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 2
      },
      "startedAt": "2026-10-19T06:24:14.205Z",
      "finishedAt": "2026-10-19T06:34:35.573Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 621368,
        "timing": {
          "startedAt": "2026-10-19T06:24:14.205Z",
          "finishedAt": "2026-10-19T06:34:35.573Z",
          "timeToFirstArtifactMs": 88281,
          "timeToLastArtifactWriteMs": 611780,
          "timeToFirstTokenMs": 33832,
          "firstTurnTtftMs": 27284,
          "timeToFirstToolCallMs": 35972
        },
        "toolCalls": 45,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.24,
        "prefillTokensPerSec": 379.62,
        "decodeTokens": 10015,
        "peakRssMb": 49642,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T06:46:26.526Z",
        "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 953,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 9,
          "numbering continuity": 9,
          "professionalism": 5,
          "completeness": 4,
          "print-polish": 8
        },
        "meanScore": 7,
        "justification": "Amounts total $5,415 and trace exactly to three uninvoiced entries; 042–044 exclude the old bill and idle client. Clean invoices each render on one page with explicit margins. Due dates wrongly start from work dates, and issue dates are absent. The ledger still says last 041 with all three entries uninvoiced; the summary’s “went out” claim lacks sending evidence.",
        "artifactSha256": "8e135c53126a5bb59cf416c9235e6e64d27c4c2bd9f58c4a6d8f91d6c6e9a56e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 2
      },
      "startedAt": "2026-10-19T06:34:46.696Z",
      "finishedAt": "2026-10-19T08:14:53.258Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 6006562,
        "timing": {
          "startedAt": "2026-10-19T06:34:46.696Z",
          "finishedAt": "2026-10-19T08:14:53.258Z",
          "timeToFirstArtifactMs": 1846462,
          "timeToLastArtifactWriteMs": 6000938,
          "timeToFirstTokenMs": 54247,
          "firstTurnTtftMs": 52119,
          "timeToFirstToolCallMs": 63491
        },
        "toolCalls": 74,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 19.04,
        "prefillTokensPerSec": 316.62,
        "decodeTokens": 43841,
        "peakRssMb": 52809,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T08:15:08.785Z",
        "trialId": "craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 2
      },
      "startedAt": "2026-10-19T08:15:06.264Z",
      "finishedAt": "2026-10-19T08:39:34.909Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-route-multi:10:targetnone:fr1mhd7ue:rp0:rf0\" stuck for 10m — 3 re-writes (10 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.1,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 90.9% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1468645,
        "timing": {
          "startedAt": "2026-10-19T08:15:06.264Z",
          "finishedAt": "2026-10-19T08:39:34.909Z",
          "timeToFirstArtifactMs": 312453,
          "timeToLastArtifactWriteMs": 1463165,
          "timeToFirstTokenMs": 39044,
          "firstTurnTtftMs": 36942,
          "timeToFirstToolCallMs": 43099
        },
        "toolCalls": 37,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.29,
        "prefillTokensPerSec": 362.68,
        "decodeTokens": 9125,
        "peakRssMb": 46848,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T08:45:08.825Z",
        "trialId": "craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 2
      },
      "startedAt": "2026-10-19T08:39:44.825Z",
      "finishedAt": "2026-10-19T09:35:55.243Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3370418,
        "timing": {
          "startedAt": "2026-10-19T08:39:44.825Z",
          "finishedAt": "2026-10-19T09:35:55.243Z",
          "timeToFirstArtifactMs": 1512053,
          "timeToLastArtifactWriteMs": 3359777,
          "timeToFirstTokenMs": 53720,
          "firstTurnTtftMs": 52144,
          "timeToFirstToolCallMs": 295830
        },
        "toolCalls": 50,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.53,
        "prefillTokensPerSec": 325.58,
        "decodeTokens": 34004,
        "peakRssMb": 54102,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T09:45:10.168Z",
        "trialId": "craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 2
      },
      "startedAt": "2026-10-19T09:36:06.747Z",
      "finishedAt": "2026-10-19T10:22:49.083Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2802336,
        "timing": {
          "startedAt": "2026-10-19T09:36:06.747Z",
          "finishedAt": "2026-10-19T10:22:49.083Z",
          "timeToFirstArtifactMs": 904965,
          "timeToLastArtifactWriteMs": 2800134,
          "timeToFirstTokenMs": 54053,
          "firstTurnTtftMs": 51943,
          "timeToFirstToolCallMs": 65384
        },
        "toolCalls": 54,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.42,
        "prefillTokensPerSec": 317.27,
        "decodeTokens": 19773,
        "peakRssMb": 52839,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T10:30:07.611Z",
        "trialId": "craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 2
      },
      "startedAt": "2026-10-19T10:23:00.275Z",
      "finishedAt": "2026-10-19T11:32:58.463Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-gate-script:4:targetnone:fr1q7wvh:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 4 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.7,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4198188,
        "timing": {
          "startedAt": "2026-10-19T10:23:00.275Z",
          "finishedAt": "2026-10-19T11:32:58.463Z",
          "timeToFirstArtifactMs": 2613021,
          "timeToLastArtifactWriteMs": 2613021,
          "timeToFirstTokenMs": 52726,
          "firstTurnTtftMs": 51120,
          "timeToFirstToolCallMs": 71648
        },
        "toolCalls": 44,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 15.39,
        "prefillTokensPerSec": 239.05,
        "decodeTokens": 26707,
        "peakRssMb": 53088,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T11:45:09.507Z",
        "trialId": "craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0",
      "cell": {
        "id": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 2
      },
      "startedAt": "2026-10-19T11:33:33.174Z",
      "finishedAt": "2026-10-19T12:07:17.484Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:6:targetnone:fr1tjwffw:rp0:rf0:m1\" stuck for 10m — 3 re-writes (9 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2024310,
        "timing": {
          "startedAt": "2026-10-19T11:33:33.174Z",
          "finishedAt": "2026-10-19T12:07:17.484Z",
          "timeToFirstArtifactMs": 761104,
          "timeToLastArtifactWriteMs": 2018699,
          "timeToFirstTokenMs": 52862,
          "firstTurnTtftMs": 50749,
          "timeToFirstToolCallMs": 71165
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.81,
        "prefillTokensPerSec": 339.46,
        "decodeTokens": 14559,
        "peakRssMb": 47814,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:15:10.751Z",
        "trialId": "craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0",
        "cellId": "r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:07:26.472Z",
      "finishedAt": "2026-10-19T12:09:56.869Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 150397,
        "timing": {
          "startedAt": "2026-10-19T12:07:26.472Z",
          "finishedAt": "2026-10-19T12:09:56.869Z",
          "timeToFirstArtifactMs": 146568,
          "timeToLastArtifactWriteMs": 146572,
          "timeToFirstTokenMs": 25823,
          "firstTurnTtftMs": 23924,
          "timeToFirstToolCallMs": 43612
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.96,
        "prefillTokensPerSec": 418.26,
        "decodeTokens": 2601,
        "peakRssMb": 40651,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:16:24.675Z",
        "trialId": "tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5",
        "cellId": "r3-qwen3.8-27b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 3580,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "Clean dark board, contrasting marks and a reset control meet the brief. Local browser checks confirm all eight X win lines, an O win, draw, occupied-cell protection and reset; the board fits 320px. Readable state handling and winning-line styling help. Empty buttons lack accessible names and changing status has no live announcement; presentation is basic.",
        "artifactSha256": "8488aa56ff0ebf989e49e574d38b3d881227478d69df668298be9c5abf41c990",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:10:32.239Z",
      "finishedAt": "2026-10-19T12:15:46.801Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 314562,
        "timing": {
          "startedAt": "2026-10-19T12:10:32.239Z",
          "finishedAt": "2026-10-19T12:15:46.801Z",
          "timeToFirstArtifactMs": 302855,
          "timeToLastArtifactWriteMs": 302861,
          "timeToFirstTokenMs": 24076,
          "firstTurnTtftMs": 22533,
          "timeToFirstToolCallMs": 44281
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 27.4,
        "prefillTokensPerSec": 439.9,
        "decodeTokens": 6044,
        "peakRssMb": 50553,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:17:47.944Z",
        "trialId": "petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q",
        "cellId": "r3-qwen3.8-27b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/workspace/pet-shop-website/index.html",
        "artifactBytes": 14294,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 6,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6,
        "justification": "Warm, readable catalog; filters, cart arithmetic and validation work locally. The generated PNG loads but is abstract rather than a pet-shop mark. Mobile header is cramped. Checkout claims an order and contact promises a reply without sending anything. HTML was first saved after image generation, violating the requested order. Cart lacks focus/Escape handling.",
        "artifactSha256": "844b6d319319c54c02e744b5c04d3c249fbda75936c2d3598a023d59565c2005",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:16:18.577Z",
      "finishedAt": "2026-10-19T12:20:50.090Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 271513,
        "timing": {
          "startedAt": "2026-10-19T12:16:18.577Z",
          "finishedAt": "2026-10-19T12:20:50.090Z",
          "timeToFirstArtifactMs": 267113,
          "timeToLastArtifactWriteMs": 267119,
          "timeToFirstTokenMs": 26037,
          "firstTurnTtftMs": 24492,
          "timeToFirstToolCallMs": 96152
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.84,
        "prefillTokensPerSec": 434.15,
        "decodeTokens": 6520,
        "peakRssMb": 39079,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:31:22.293Z",
        "trialId": "tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9",
        "cellId": "r3-qwen3.8-27b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 12162,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 6
        },
        "meanScore": 6.8,
        "justification": "Readable grid arena, score/lives/wave HUD and controls guide. Local browser checks confirm movement, spawning, shot scoring, enemy damage, game over and restart. Delta-time motion and normalized diagonals are sound; particles and invulnerability flashing add feedback. The fixed 800px canvas overflows narrow screens, and focus loss can leave movement keys held.",
        "artifactSha256": "e02171d1b57299b6dbc227c2869056b9f7118e76a6fd73f7ea6ceff2067a6d38",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:21:25.640Z",
      "finishedAt": "2026-10-19T12:25:39.698Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 254058,
        "timing": {
          "startedAt": "2026-10-19T12:21:25.640Z",
          "finishedAt": "2026-10-19T12:25:39.698Z",
          "timeToFirstArtifactMs": 23021,
          "timeToLastArtifactWriteMs": 250423,
          "timeToFirstTokenMs": 13868,
          "firstTurnTtftMs": 12264,
          "timeToFirstToolCallMs": 23023
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 29.68,
        "prefillTokensPerSec": 438.43,
        "decodeTokens": 3247,
        "peakRssMb": 48799,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:30:09.861Z",
        "trialId": "schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe",
        "cellId": "r3-qwen3.8-27b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:25:48.560Z",
      "finishedAt": "2026-10-19T12:26:41.656Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 53096,
        "timing": {
          "startedAt": "2026-10-19T12:25:48.560Z",
          "finishedAt": "2026-10-19T12:26:41.656Z",
          "timeToFirstArtifactMs": 51036,
          "timeToLastArtifactWriteMs": 51038,
          "timeToFirstTokenMs": 18024,
          "firstTurnTtftMs": 16405,
          "timeToFirstToolCallMs": 19457
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 29.64,
        "prefillTokensPerSec": 393.47,
        "decodeTokens": 618,
        "peakRssMb": 37826,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:30:09.861Z",
        "trialId": "failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny",
        "cellId": "r3-qwen3.8-27b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:26:50.565Z",
      "finishedAt": "2026-10-19T12:27:53.096Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 62531,
        "timing": {
          "startedAt": "2026-10-19T12:26:50.565Z",
          "finishedAt": "2026-10-19T12:27:53.096Z",
          "timeToFirstArtifactMs": 58281,
          "timeToLastArtifactWriteMs": 58281,
          "timeToFirstTokenMs": 18113,
          "firstTurnTtftMs": 16520,
          "timeToFirstToolCallMs": 20139
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.36,
        "prefillTokensPerSec": 402.48,
        "decodeTokens": 778,
        "peakRssMb": 37205,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:30:09.861Z",
        "trialId": "symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b",
        "cellId": "r3-qwen3.8-27b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:28:02.592Z",
      "finishedAt": "2026-10-19T12:30:45.773Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 163181,
        "timing": {
          "startedAt": "2026-10-19T12:28:02.592Z",
          "finishedAt": "2026-10-19T12:30:45.773Z",
          "timeToFirstArtifactMs": 65203,
          "timeToLastArtifactWriteMs": 144616,
          "timeToFirstTokenMs": 18198,
          "firstTurnTtftMs": 16610,
          "timeToFirstToolCallMs": 20887
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.42,
        "prefillTokensPerSec": 390.64,
        "decodeTokens": 3661,
        "peakRssMb": 38042,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T12:31:22.551Z",
        "trialId": "data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv",
        "cellId": "r3-qwen3.8-27b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:30:55.156Z",
      "finishedAt": "2026-10-19T12:46:02.287Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 907131,
        "timing": {
          "startedAt": "2026-10-19T12:30:55.156Z",
          "finishedAt": "2026-10-19T12:46:02.287Z",
          "timeToFirstArtifactMs": 900811,
          "timeToLastArtifactWriteMs": 900811,
          "timeToFirstTokenMs": 36990,
          "firstTurnTtftMs": 35272,
          "timeToFirstToolCallMs": 44403
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.76,
        "prefillTokensPerSec": 360.5,
        "decodeTokens": 17573,
        "peakRssMb": 44896,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:00:10.859Z",
        "trialId": "incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl",
        "cellId": "r3-qwen3.8-27b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:46:11.942Z",
      "finishedAt": "2026-10-19T12:48:19.709Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 127767,
        "timing": {
          "startedAt": "2026-10-19T12:46:11.942Z",
          "finishedAt": "2026-10-19T12:48:19.709Z",
          "timeToFirstArtifactMs": 50847,
          "timeToLastArtifactWriteMs": 125028,
          "timeToFirstTokenMs": 35984,
          "firstTurnTtftMs": 34389,
          "timeToFirstToolCallMs": 37926
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.31,
        "prefillTokensPerSec": 397.8,
        "decodeTokens": 1445,
        "peakRssMb": 41790,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:00:10.860Z",
        "trialId": "ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha",
        "cellId": "r3-qwen3.8-27b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:48:29.306Z",
      "finishedAt": "2026-10-19T12:50:12.180Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 102874,
        "timing": {
          "startedAt": "2026-10-19T12:48:29.306Z",
          "finishedAt": "2026-10-19T12:50:12.180Z",
          "timeToFirstArtifactMs": 100022,
          "timeToLastArtifactWriteMs": 100022,
          "timeToFirstTokenMs": 18871,
          "firstTurnTtftMs": 17181,
          "timeToFirstToolCallMs": 22060
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 18.21,
        "prefillTokensPerSec": 408.35,
        "decodeTokens": 1454,
        "peakRssMb": 37025,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:00:10.860Z",
        "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad",
        "cellId": "r3-qwen3.8-27b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-19T12:50:21.596Z",
      "finishedAt": "2026-10-19T13:00:36.887Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 615291,
        "timing": {
          "startedAt": "2026-10-19T12:50:21.596Z",
          "finishedAt": "2026-10-19T13:00:36.887Z",
          "timeToFirstArtifactMs": 295878,
          "timeToLastArtifactWriteMs": 611317,
          "timeToFirstTokenMs": 34258,
          "firstTurnTtftMs": 32655,
          "timeToFirstToolCallMs": 40928
        },
        "toolCalls": 27,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.41,
        "prefillTokensPerSec": 390.27,
        "decodeTokens": 8581,
        "peakRssMb": 48827,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:15:10.206Z",
        "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8",
        "cellId": "r3-qwen3.8-27b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:00:45.988Z",
      "finishedAt": "2026-10-19T13:01:48.477Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 62489,
        "timing": {
          "startedAt": "2026-10-19T13:00:45.988Z",
          "finishedAt": "2026-10-19T13:01:48.477Z",
          "timeToFirstArtifactMs": 23333,
          "timeToLastArtifactWriteMs": 61222,
          "timeToFirstTokenMs": 13752,
          "firstTurnTtftMs": 12154,
          "timeToFirstToolCallMs": 23335
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.65,
        "prefillTokensPerSec": 425.19,
        "decodeTokens": 658,
        "peakRssMb": 38591,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:15:10.206Z",
        "trialId": "constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df",
        "cellId": "r3-qwen3.8-27b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:01:58.058Z",
      "finishedAt": "2026-10-19T13:13:33.633Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 695575,
        "timing": {
          "startedAt": "2026-10-19T13:01:58.058Z",
          "finishedAt": "2026-10-19T13:13:33.633Z",
          "timeToFirstArtifactMs": 276008,
          "timeToLastArtifactWriteMs": 692018,
          "timeToFirstTokenMs": 18702,
          "firstTurnTtftMs": 17114,
          "timeToFirstToolCallMs": 20422
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.32,
        "prefillTokensPerSec": 376.46,
        "decodeTokens": 16053,
        "peakRssMb": 40493,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:15:10.206Z",
        "trialId": "records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em",
        "cellId": "r3-qwen3.8-27b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:13:45.318Z",
      "finishedAt": "2026-10-19T13:16:13.172Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 147854,
        "timing": {
          "startedAt": "2026-10-19T13:13:45.318Z",
          "finishedAt": "2026-10-19T13:16:13.172Z",
          "timeToFirstArtifactMs": 145660,
          "timeToLastArtifactWriteMs": 145660,
          "timeToFirstTokenMs": 19293,
          "firstTurnTtftMs": 17672,
          "timeToFirstToolCallMs": 145662
        },
        "toolCalls": 1,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.39,
        "prefillTokensPerSec": 398.55,
        "decodeTokens": 2197,
        "peakRssMb": 36952,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:30:10.143Z",
        "trialId": "plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z",
        "cellId": "r3-qwen3.8-27b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:16:22.825Z",
      "finishedAt": "2026-10-19T13:19:01.151Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 158326,
        "timing": {
          "startedAt": "2026-10-19T13:16:22.825Z",
          "finishedAt": "2026-10-19T13:19:01.151Z",
          "timeToFirstArtifactMs": 52388,
          "timeToLastArtifactWriteMs": 156255,
          "timeToFirstTokenMs": 26134,
          "firstTurnTtftMs": 24540,
          "timeToFirstToolCallMs": 28390
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 33.72,
        "prefillTokensPerSec": 430.68,
        "decodeTokens": 1015,
        "peakRssMb": 38786,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:33:21.347Z",
        "trialId": "meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e",
        "cellId": "r3-qwen3.8-27b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1509,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 8,
          "grounding": 9,
          "actionability": 9,
          "brevity": 8
        },
        "meanScore": 8.5,
        "justification": "Clear decisions and exact four-row CSV preserve owners, deadlines and A2→A1 dependency; stale proposals stay rejected and legal/rollback gaps remain open. The open-question prose says Jordan will run the review by Aug 30, while the transcript only promises scheduling it by then. Some stale-proposal explanations repeat.",
        "artifactSha256": "203706c9897372fdfe38565e714cb1b8061b1c357c531a3eb014d47e42aa6d1f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:19:10.699Z",
      "finishedAt": "2026-10-19T13:29:57.789Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 647090,
        "timing": {
          "startedAt": "2026-10-19T13:19:10.699Z",
          "finishedAt": "2026-10-19T13:29:57.789Z",
          "timeToFirstArtifactMs": 107196,
          "timeToLastArtifactWriteMs": 427611,
          "timeToFirstTokenMs": 25456,
          "firstTurnTtftMs": 18949,
          "timeToFirstToolCallMs": 29697
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.82,
        "prefillTokensPerSec": 438.7,
        "decodeTokens": 12816,
        "peakRssMb": 43699,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:33:21.355Z",
        "trialId": "craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9",
        "cellId": "r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 2664,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 9,
          "usefulness": 7
        },
        "meanScore": 8,
        "justification": "Scope precedes build and reviewer hand-checks defaults and sanity scenario before finish; live inputs/reset reproduce $627.36/26.1% and $40/20%. Clear calculation zones. Scope misstates C9 as $1,427.36 (runtime $1,427.04); monthly revenue and applying leakage to all revenue are unsupported assumptions presented as source facts.",
        "artifactSha256": "b80b6c686dd40ae5f7d02fe6592e342ed0bc407d71978885c4f2993d83544ae2",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:30:07.253Z",
      "finishedAt": "2026-10-19T13:36:21.103Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 373850,
        "timing": {
          "startedAt": "2026-10-19T13:30:07.253Z",
          "finishedAt": "2026-10-19T13:36:21.103Z",
          "timeToFirstArtifactMs": 213036,
          "timeToLastArtifactWriteMs": 367822,
          "timeToFirstTokenMs": 35446,
          "firstTurnTtftMs": 33851,
          "timeToFirstToolCallMs": 42434
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 26.05,
        "prefillTokensPerSec": 388.41,
        "decodeTokens": 5587,
        "peakRssMb": 47518,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:45:07.484Z",
        "trialId": "conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh",
        "cellId": "r3-qwen3.8-27b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:36:30.696Z",
      "finishedAt": "2026-10-19T13:38:49.024Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 138328,
        "timing": {
          "startedAt": "2026-10-19T13:36:30.696Z",
          "finishedAt": "2026-10-19T13:38:49.024Z",
          "timeToFirstArtifactMs": 135240,
          "timeToLastArtifactWriteMs": 135240,
          "timeToFirstTokenMs": 22788,
          "firstTurnTtftMs": 21111,
          "timeToFirstToolCallMs": 26503
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.45,
        "prefillTokensPerSec": 375.79,
        "decodeTokens": 2164,
        "peakRssMb": 44279,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T13:45:38.130Z",
        "trialId": "docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj",
        "cellId": "r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 2153,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 9,
          "candor": 9,
          "usefulness": 8
        },
        "meanScore": 8.7,
        "justification": "Theme id, fonts, palette, 11→9 pages, applied styles and content-preserved claim match captured DocBlocks responses. Callout Emphasis is plainly unresolved and retains source formatting. Conditional advice to accept an exception or ask the brand team for a variant is actionable; the extra per-style uncertainty is cautious but slightly confusing.",
        "artifactSha256": "1c817d828e221fd9d0f45eeb4e953d15c68eacbe4813a11c19181082f9d6dd85",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:39:00.372Z",
      "finishedAt": "2026-10-19T13:48:36.042Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 575670,
        "timing": {
          "startedAt": "2026-10-19T13:39:00.372Z",
          "finishedAt": "2026-10-19T13:48:36.042Z",
          "timeToFirstArtifactMs": 102591,
          "timeToLastArtifactWriteMs": 572087,
          "timeToFirstTokenMs": 38309,
          "firstTurnTtftMs": 31688,
          "timeToFirstToolCallMs": 47403
        },
        "toolCalls": 31,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.54,
        "prefillTokensPerSec": 389.47,
        "decodeTokens": 8795,
        "peakRssMb": 49652,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T14:00:43.680Z",
        "trialId": "craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6",
        "cellId": "r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 2277,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 7,
          "convertibility": 8
        },
        "meanScore": 8.3,
        "justification": "Exact supplied brief was read and cited before outlining; seven H1s match all outline titles in order. All metrics and three next actions survive unchanged, and DocBlocks conversion/preview/save produced matching PPTX copies. “Demonstrably solved” overstates 18h→6h without a target; process-versus-tooling conclusions are unsupported despite the proposal label.",
        "artifactSha256": "247eddbf1d4e30c53babcbbcdf575b940c59da179295e54041e68cbb4d859c6e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 3
      },
      "startedAt": "2026-10-19T13:48:48.025Z",
      "finishedAt": "2026-10-19T14:00:53.692Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 725667,
        "timing": {
          "startedAt": "2026-10-19T13:48:48.025Z",
          "finishedAt": "2026-10-19T14:00:53.692Z",
          "timeToFirstArtifactMs": 641584,
          "timeToLastArtifactWriteMs": 719817,
          "timeToFirstTokenMs": 15707,
          "firstTurnTtftMs": 14041,
          "timeToFirstToolCallMs": 99590
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.51,
        "prefillTokensPerSec": 380.64,
        "decodeTokens": 13491,
        "peakRssMb": 42143,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T14:15:31.926Z",
        "trialId": "wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe",
        "cellId": "r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5452,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 6,
          "grounding": 7,
          "citationQuality": 8,
          "calibration": 8
        },
        "meanScore": 7.3,
        "justification": "Correct chronology and engine architecture are linked to all three relevant cards, with nearby ids and explicit debate around the programmer label. Much of the brief repeats source summaries. “Never ran,” exclusive contribution assignments and the algorithm-versus-closed-form distinction go beyond the cards; lifetime non-completion alone does not establish those broader claims.",
        "artifactSha256": "6973c99c07b3a81c95fed7fbe946d3172ccc19c9ee9b001245d5c904eeba4e60",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 3
      },
      "startedAt": "2026-10-19T14:01:03.377Z",
      "finishedAt": "2026-10-19T14:32:13.738Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-week-plan passed 12 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1870361,
        "timing": {
          "startedAt": "2026-10-19T14:01:03.377Z",
          "finishedAt": "2026-10-19T14:32:13.738Z",
          "timeToFirstArtifactMs": 184928,
          "timeToLastArtifactWriteMs": 1797017,
          "timeToFirstTokenMs": 47612,
          "firstTurnTtftMs": 41027,
          "timeToFirstToolCallMs": 52872
        },
        "toolCalls": 74,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.93,
        "prefillTokensPerSec": 396.1,
        "decodeTokens": 29878,
        "peakRssMb": 50395,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T14:45:50.424Z",
        "trialId": "craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0",
        "cellId": "r3-qwen3.8-27b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/workspace/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 6480,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 1,
          "prioritization": 1,
          "conflict-handling": 0,
          "realism": 1
        },
        "meanScore": 0.8,
        "justification": "The probe finally fetched the July 13 eight-event fixture, but the report still invents an October week, Sam/Meridian/onboarding meetings and daily standups. It misses the real Tuesday budget/Priya overlap. Thursday focus overlaps Design review and Friday focus overlaps retro; SOW work follows its invented demo. Honest-load claims are unsupported.",
        "artifactSha256": "030cc933cfdb2fa250a29ef9c17a9172b8681b7812c5f89eeda67bb0de71f474",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 3
      },
      "startedAt": "2026-10-19T14:32:25.886Z",
      "finishedAt": "2026-10-19T15:03:26.229Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1860343,
        "timing": {
          "startedAt": "2026-10-19T14:32:25.886Z",
          "finishedAt": "2026-10-19T15:03:26.229Z",
          "timeToFirstArtifactMs": 168903,
          "timeToLastArtifactWriteMs": 1854937,
          "timeToFirstTokenMs": 45325,
          "firstTurnTtftMs": 38779,
          "timeToFirstToolCallMs": 71400
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 24.35,
        "prefillTokensPerSec": 377.49,
        "decodeTokens": 33045,
        "peakRssMb": 51406,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:15:07.690Z",
        "trialId": "craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x",
        "cellId": "r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2247,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:03:37.951Z",
      "finishedAt": "2026-10-19T15:17:30.071Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 832120,
        "timing": {
          "startedAt": "2026-10-19T15:03:37.951Z",
          "finishedAt": "2026-10-19T15:17:30.071Z",
          "timeToFirstArtifactMs": 94089,
          "timeToLastArtifactWriteMs": 667358,
          "timeToFirstTokenMs": 31875,
          "firstTurnTtftMs": 25337,
          "timeToFirstToolCallMs": 94088
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.58,
        "prefillTokensPerSec": 391.83,
        "decodeTokens": 15326,
        "peakRssMb": 46688,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:30:47.654Z",
        "trialId": "craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql",
        "cellId": "r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 6809,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 8,
          "grounding": 5,
          "structure": 8,
          "actionability": 7
        },
        "meanScore": 7,
        "justification": "All six sources and their numerical findings are retained in uniform APA-style entries with three labeled annotations. Critical limitations are often sensibly hedged. But Chen is called cross-sectional without evidence, connectivity becomes a primary cause, and Patel’s break-even model is treated as proof of viability at realistic rural volumes; these exceed the cards.",
        "artifactSha256": "d57ade2e20751d8523915098a1290dd69dae92d01f94907905196a10a296da1c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:17:39.427Z",
      "finishedAt": "2026-10-19T15:25:56.871Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 497444,
        "timing": {
          "startedAt": "2026-10-19T15:17:39.427Z",
          "finishedAt": "2026-10-19T15:25:56.871Z",
          "timeToFirstArtifactMs": 69600,
          "timeToLastArtifactWriteMs": 309048,
          "timeToFirstTokenMs": 31396,
          "firstTurnTtftMs": 24845,
          "timeToFirstToolCallMs": 35747
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23,
        "prefillTokensPerSec": 393.44,
        "decodeTokens": 7382,
        "peakRssMb": 49461,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:30:47.656Z",
        "trialId": "craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j",
        "cellId": "r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1815,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 7,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 7.7,
        "justification": "Compact title, summary, metrics table, findings and three next actions form complete DOCX-ready Markdown; DocBlocks conversion, preview and artifact save are captured. Supplied figures are correct, including valid derived comparisons. “The one complaint” overstates “top unresolved,” while barcode exceptions stalling returns and actions already being in place are unsupported.",
        "artifactSha256": "0ff041b1ada0fe40e6deffd7c80f97086a7d318f11750dde0c747731d88ef13f",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:26:06.425Z",
      "finishedAt": "2026-10-19T15:29:04.695Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 178270,
        "timing": {
          "startedAt": "2026-10-19T15:26:06.425Z",
          "finishedAt": "2026-10-19T15:29:04.695Z",
          "timeToFirstArtifactMs": 91313,
          "timeToLastArtifactWriteMs": 172039,
          "timeToFirstTokenMs": 17754,
          "firstTurnTtftMs": 16147,
          "timeToFirstToolCallMs": 22082
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 28.24,
        "prefillTokensPerSec": 389.09,
        "decodeTokens": 3932,
        "peakRssMb": 38388,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:30:07.722Z",
        "trialId": "fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839",
        "cellId": "r3-qwen3.8-27b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:29:13.944Z",
      "finishedAt": "2026-10-19T15:30:32.196Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 78252,
        "timing": {
          "startedAt": "2026-10-19T15:29:13.944Z",
          "finishedAt": "2026-10-19T15:30:32.196Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 37790,
          "firstTurnTtftMs": 36220,
          "timeToFirstToolCallMs": 48992
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 21.34,
        "prefillTokensPerSec": 399.24,
        "decodeTokens": 675,
        "peakRssMb": 37537,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:30:47.902Z",
        "trialId": "dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2",
        "cellId": "r3-qwen3.8-27b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:30:41.095Z",
      "finishedAt": "2026-10-19T15:35:21.466Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 5 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 280371,
        "timing": {
          "startedAt": "2026-10-19T15:30:41.095Z",
          "finishedAt": "2026-10-19T15:35:21.466Z",
          "timeToFirstArtifactMs": 274678,
          "timeToLastArtifactWriteMs": 274698,
          "timeToFirstTokenMs": 25932,
          "firstTurnTtftMs": 24377,
          "timeToFirstToolCallMs": 131976
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.58,
        "prefillTokensPerSec": 399.57,
        "decodeTokens": 4523,
        "peakRssMb": 41225,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:45:08.099Z",
        "trialId": "interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175",
        "cellId": "r3-qwen3.8-27b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:35:31.027Z",
      "finishedAt": "2026-10-19T15:42:03.977Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 392950,
        "timing": {
          "startedAt": "2026-10-19T15:35:31.027Z",
          "finishedAt": "2026-10-19T15:42:03.977Z",
          "timeToFirstArtifactMs": 289419,
          "timeToLastArtifactWriteMs": 289419,
          "timeToFirstTokenMs": 30549,
          "firstTurnTtftMs": 24021,
          "timeToFirstToolCallMs": 32847
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.51,
        "prefillTokensPerSec": 398.05,
        "decodeTokens": 5287,
        "peakRssMb": 45434,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T15:45:43.519Z",
        "trialId": "craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge",
        "cellId": "r3-qwen3.8-27b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3655,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 7
        },
        "meanScore": 7.5,
        "justification": "Scope→report→done uses the snapshot and conventions, saves/re-reads the artifact and stamps request-changes. Both SQL regressions, handle leak, false-success catch, pagination and debug log are found; deliberate safe patterns stay clear. Payment citations are inaccurate: SELECT is line 7, INSERT 11 and debug log 9, versus reported 9/12/11.",
        "artifactSha256": "fd7d764042d3f0a11f3fe4d6f9f796b2bb3dcae93e17b37e9cfe0ab57876ea7b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 3
      },
      "startedAt": "2026-10-19T15:42:13.494Z",
      "finishedAt": "2026-10-19T16:15:10.324Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1976830,
        "timing": {
          "startedAt": "2026-10-19T15:42:13.494Z",
          "finishedAt": "2026-10-19T16:15:10.324Z",
          "timeToFirstArtifactMs": 561875,
          "timeToLastArtifactWriteMs": 1848719,
          "timeToFirstTokenMs": 33115,
          "firstTurnTtftMs": 26598,
          "timeToFirstToolCallMs": 54216
        },
        "toolCalls": 28,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.76,
        "prefillTokensPerSec": 369.57,
        "decodeTokens": 35598,
        "peakRssMb": 46049,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T16:15:48.388Z",
        "trialId": "craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj",
        "cellId": "r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 12968,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 5
        },
        "meanScore": 6.5,
        "justification": "Scope/checklist, endpoint review, final report and evaluation follow the craftbook. The three real drifts are found, and imported sessions plus optional nullable email are correctly cleared. But v1.0.0 does not prove no existing consumers, so “Backward-compatible: yes” is unsupported. Collapsed code fences and YAML indentation make most corrected snippets unusable.",
        "artifactSha256": "383525bca0463f06c58fb0d045a59fe29b0c913bb472a0d9d759e7d4b63581ba",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 3
      },
      "startedAt": "2026-10-19T16:15:19.947Z",
      "finishedAt": "2026-10-19T16:21:26.643Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "craftbook-deep-security-review stale no-write repair loop: security-review/REPORT.md had unchanged content (4850 bytes) for 71s — a re-emission of identical bytes counts as unchanged here after repair feedback and newer chat activity; still failing 9/11: security-review/REPORT.md is missing required content /Verified Safe|Not Statically Verifiable/ — nothing in its 4850 bytes matches. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.8,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 81.8% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 366696,
        "timing": {
          "startedAt": "2026-10-19T16:15:19.947Z",
          "finishedAt": "2026-10-19T16:21:26.643Z",
          "timeToFirstArtifactMs": 156138,
          "timeToLastArtifactWriteMs": 323699,
          "timeToFirstTokenMs": 30251,
          "firstTurnTtftMs": 28647,
          "timeToFirstToolCallMs": 35830
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 25.15,
        "prefillTokensPerSec": 376.8,
        "decodeTokens": 6041,
        "peakRssMb": 46724,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T16:30:38.598Z",
        "trialId": "craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1",
        "cellId": "r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 4872,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 7,
          "remediation": 8
        },
        "meanScore": 7.7,
        "justification": "All three admin findings trace request inputs to shell/SQL sinks or absent authorization, with exact lines and severity-consistent JSON fixes. Profile, static shell and allow-listed SQL are correctly cleared. Root causes and blast radius are grouped, but claims that no enforcement exists and SQL permits arbitrary writes exceed static evidence.",
        "artifactSha256": "e7b24a363bcda6f078dbe627c498a3a2ac1694841f8642a57d07cff7ff92bd75",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 3
      },
      "startedAt": "2026-10-19T16:21:36.254Z",
      "finishedAt": "2026-10-19T16:58:52.183Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2235929,
        "timing": {
          "startedAt": "2026-10-19T16:21:36.254Z",
          "finishedAt": "2026-10-19T16:58:52.183Z",
          "timeToFirstArtifactMs": 1850046,
          "timeToLastArtifactWriteMs": 2230183,
          "timeToFirstTokenMs": 42953,
          "firstTurnTtftMs": 41238,
          "timeToFirstToolCallMs": 46117
        },
        "toolCalls": 39,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 20.27,
        "prefillTokensPerSec": 234.61,
        "decodeTokens": 29108,
        "peakRssMb": 53668,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T17:00:15.225Z",
        "trialId": "large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172",
        "cellId": "r3-qwen3.8-27b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 3
      },
      "startedAt": "2026-10-19T16:59:01.589Z",
      "finishedAt": "2026-10-19T17:09:26.022Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 624433,
        "timing": {
          "startedAt": "2026-10-19T16:59:01.589Z",
          "finishedAt": "2026-10-19T17:09:26.022Z",
          "timeToFirstArtifactMs": 138642,
          "timeToLastArtifactWriteMs": 561161,
          "timeToFirstTokenMs": 32294,
          "firstTurnTtftMs": 25737,
          "timeToFirstToolCallMs": 37047
        },
        "toolCalls": 38,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 23.32,
        "prefillTokensPerSec": 394.19,
        "decodeTokens": 9750,
        "peakRssMb": 48957,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T17:15:45.290Z",
        "trialId": "craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk",
        "cellId": "r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 4176,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 9,
          "Invariant discipline": 9,
          "Reconciled verification": 9
        },
        "meanScore": 9,
        "justification": "Real pre-edit ripgrep found all 11 matches before the first replacement: seven executable/log sites and four justified history/spec exclusions. Captured edits are token-only, including re-export and template log; exclusions stay intact. Post-edit grep and full-file verification reconcile 7 changed/4 excluded/0 unwanted residuals, with two real 3/3 test runs.",
        "artifactSha256": "134a69a5670a98057313cfdb2b317cb938c502e768529ab59dc071f6792afc0a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 3
      },
      "startedAt": "2026-10-19T17:09:35.000Z",
      "finishedAt": "2026-10-19T17:44:12.443Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2077443,
        "timing": {
          "startedAt": "2026-10-19T17:09:35.000Z",
          "finishedAt": "2026-10-19T17:44:12.443Z",
          "timeToFirstArtifactMs": 459514,
          "timeToLastArtifactWriteMs": 1993075,
          "timeToFirstTokenMs": 25290,
          "firstTurnTtftMs": 18762,
          "timeToFirstToolCallMs": 51553
        },
        "toolCalls": 70,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.05,
        "prefillTokensPerSec": 373.43,
        "decodeTokens": 38200,
        "peakRssMb": 52408,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T17:45:47.597Z",
        "trialId": "craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9",
        "cellId": "r3-qwen3.8-27b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 2466,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 9,
          "Reversible-stage execution": 9,
          "Genuine deduplication": 9
        },
        "meanScore": 9,
        "justification": "Captured 2/2 baseline and 12/12 characterization runs precede source edits. Three planned stages each have real 12/12 green receipts; the refused rewrite/focused-edit fallback is disclosed. One unchanged algorithm serves both public call sites, preserving coercion and empty inputs. Reviewer notes the minor 11-versus-9-line documentation error.",
        "artifactSha256": "591f8b8454b3b8207c1b5103f858bdf7d1bbf4df0d620cbeeec172564faafd8b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 3
      },
      "startedAt": "2026-10-19T17:44:21.871Z",
      "finishedAt": "2026-10-19T17:57:57.310Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 815439,
        "timing": {
          "startedAt": "2026-10-19T17:44:21.871Z",
          "finishedAt": "2026-10-19T17:57:57.310Z",
          "timeToFirstArtifactMs": 51045,
          "timeToLastArtifactWriteMs": 734440,
          "timeToFirstTokenMs": 35273,
          "firstTurnTtftMs": 28725,
          "timeToFirstToolCallMs": 39854
        },
        "toolCalls": 39,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.59,
        "prefillTokensPerSec": 384.38,
        "decodeTokens": 13521,
        "peakRssMb": 48530,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T18:00:41.896Z",
        "trialId": "craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc",
        "cellId": "r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 3334,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 9,
          "Honest verification": 8
        },
        "meanScore": 8.7,
        "justification": "Real red run at 17:45:24 shows 80≠68 before the 17:48:47 one-character >= fix. Original regression survives; added 9/11 boundary checks and a real 5/5 green run verify both sides. Fix notes candidly defer execution to validation, but validation confusingly claims no source/test changes between red and green despite the fix and added test.",
        "artifactSha256": "cead9eb83b8f8d05be68e18431cede77221017ecd4f06be8e398ea551f5ff8c1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 3
      },
      "startedAt": "2026-10-19T17:58:07.088Z",
      "finishedAt": "2026-10-19T17:59:18.237Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 71149,
        "timing": {
          "startedAt": "2026-10-19T17:58:07.088Z",
          "finishedAt": "2026-10-19T17:59:18.237Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 40628,
          "firstTurnTtftMs": 39036,
          "timeToFirstToolCallMs": 46256
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 22.19,
        "prefillTokensPerSec": 370.93,
        "decodeTokens": 408,
        "peakRssMb": 37531,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T18:00:10.181Z",
        "trialId": "craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 3
      },
      "startedAt": "2026-10-19T17:59:27.100Z",
      "finishedAt": "2026-10-19T19:01:10.291Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3703191,
        "timing": {
          "startedAt": "2026-10-19T17:59:27.100Z",
          "finishedAt": "2026-10-19T19:01:10.291Z",
          "timeToFirstArtifactMs": 2457012,
          "timeToLastArtifactWriteMs": 3089805,
          "timeToFirstTokenMs": 53279,
          "firstTurnTtftMs": 51636,
          "timeToFirstToolCallMs": 59525
        },
        "toolCalls": 56,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 15.26,
        "prefillTokensPerSec": 296.37,
        "decodeTokens": 37253,
        "peakRssMb": 46993,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T19:15:08.389Z",
        "trialId": "craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 3
      },
      "startedAt": "2026-10-19T19:01:19.897Z",
      "finishedAt": "2026-10-19T19:14:17.851Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 777954,
        "timing": {
          "startedAt": "2026-10-19T19:01:19.897Z",
          "finishedAt": "2026-10-19T19:14:17.851Z",
          "timeToFirstArtifactMs": 100845,
          "timeToLastArtifactWriteMs": 765246,
          "timeToFirstTokenMs": 33963,
          "firstTurnTtftMs": 27419,
          "timeToFirstToolCallMs": 36165
        },
        "toolCalls": 52,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 26.23,
        "prefillTokensPerSec": 363.87,
        "decodeTokens": 12063,
        "peakRssMb": 51071,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T19:16:32.193Z",
        "trialId": "craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 1200,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 8,
          "numbering continuity": 9,
          "professionalism": 6,
          "completeness": 4,
          "print-polish": 8
        },
        "meanScore": 7,
        "justification": "Amounts 1840/2600/975 total 5415; 042–044 follow 041 without rebilling 1200. Due dates correctly apply net terms to Oct 19. Report uses EUR, first invoice USD, others no currency; July work is mislabeled October. All print on one clean page. Idle client skipped, but ledger still says 041 and all three entries uninvoiced; local files do not prove invoices went out.",
        "artifactSha256": "e539ef2374c57c62b7dc7520d48c835402378b459b94648113628cf0bb9326ae",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 3
      },
      "startedAt": "2026-10-19T19:14:27.074Z",
      "finishedAt": "2026-10-19T21:04:49.299Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 6622225,
        "timing": {
          "startedAt": "2026-10-19T19:14:27.074Z",
          "finishedAt": "2026-10-19T21:04:49.299Z",
          "timeToFirstArtifactMs": 1245298,
          "timeToLastArtifactWriteMs": 6316673,
          "timeToFirstTokenMs": 56105,
          "firstTurnTtftMs": 54023,
          "timeToFirstToolCallMs": 77658
        },
        "toolCalls": 81,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 17.36,
        "prefillTokensPerSec": 311.2,
        "decodeTokens": 47768,
        "peakRssMb": 52549,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T21:15:12.566Z",
        "trialId": "craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 3
      },
      "startedAt": "2026-10-19T21:05:02.403Z",
      "finishedAt": "2026-10-19T21:49:53.940Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2691537,
        "timing": {
          "startedAt": "2026-10-19T21:05:02.403Z",
          "finishedAt": "2026-10-19T21:49:53.940Z",
          "timeToFirstArtifactMs": 329324,
          "timeToLastArtifactWriteMs": 1530952,
          "timeToFirstTokenMs": 40062,
          "firstTurnTtftMs": 37976,
          "timeToFirstToolCallMs": 43611
        },
        "toolCalls": 58,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 18.5,
        "prefillTokensPerSec": 335.5,
        "decodeTokens": 16003,
        "peakRssMb": 52138,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T22:00:09.188Z",
        "trialId": "craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 3
      },
      "startedAt": "2026-10-19T21:50:03.988Z",
      "finishedAt": "2026-10-19T22:32:44.253Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2560265,
        "timing": {
          "startedAt": "2026-10-19T21:50:03.988Z",
          "finishedAt": "2026-10-19T22:32:44.253Z",
          "timeToFirstArtifactMs": 889329,
          "timeToLastArtifactWriteMs": 2554419,
          "timeToFirstTokenMs": 58178,
          "firstTurnTtftMs": 56594,
          "timeToFirstToolCallMs": 109686
        },
        "toolCalls": 41,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 18.28,
        "prefillTokensPerSec": 314.91,
        "decodeTokens": 24730,
        "peakRssMb": 47890,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T22:45:11.295Z",
        "trialId": "craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 3
      },
      "startedAt": "2026-10-19T22:32:56.184Z",
      "finishedAt": "2026-10-19T23:29:44.816Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3408632,
        "timing": {
          "startedAt": "2026-10-19T22:32:56.184Z",
          "finishedAt": "2026-10-19T23:29:44.816Z",
          "timeToFirstArtifactMs": 2045518,
          "timeToLastArtifactWriteMs": 3402504,
          "timeToFirstTokenMs": 58045,
          "firstTurnTtftMs": 55819,
          "timeToFirstToolCallMs": 78645
        },
        "toolCalls": 50,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 16.94,
        "prefillTokensPerSec": 295.5,
        "decodeTokens": 17897,
        "peakRssMb": 49320,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-19T23:30:10.594Z",
        "trialId": "craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 3
      },
      "startedAt": "2026-10-19T23:29:54.614Z",
      "finishedAt": "2026-10-19T23:51:11.987Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 603s — model busy but not delivering (hard digest stuck at c14ad1c30be188ec)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_NO_PRODUCED_OUTPUT",
            "summary": "No authored artifact event or scored sniff output was observed."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1277373,
        "timing": {
          "startedAt": "2026-10-19T23:29:54.614Z",
          "finishedAt": "2026-10-19T23:51:11.987Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 59324,
          "firstTurnTtftMs": 57742,
          "timeToFirstToolCallMs": 66086
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 12.78,
        "prefillTokensPerSec": 231.27,
        "decodeTokens": 3733,
        "peakRssMb": 44010,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T00:00:10.042Z",
        "trialId": "craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7",
      "cell": {
        "id": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "qwen3.8-27b-q4",
        "targetId": "qwen3.8-27b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 3
      },
      "startedAt": "2026-10-19T23:51:24.060Z",
      "finishedAt": "2026-10-20T00:48:28.420Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-fanout:3:targetnone:fr1tjwffw:rp0:rf0:m1\" stuck for 18m — artifact scored but stalled 18m despite 5 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 3.8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 37.5% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3424360,
        "timing": {
          "startedAt": "2026-10-19T23:51:24.060Z",
          "finishedAt": "2026-10-20T00:48:28.420Z",
          "timeToFirstArtifactMs": 979337,
          "timeToLastArtifactWriteMs": 2335820,
          "timeToFirstTokenMs": 61064,
          "firstTurnTtftMs": 58957,
          "timeToFirstToolCallMs": 88787
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "qwen3.8-27b-q4"
        },
        "decodeTokensPerSec": 12,
        "prefillTokensPerSec": 275.35,
        "decodeTokens": 18015,
        "peakRssMb": 44401,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:00:12.381Z",
        "trialId": "craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7",
        "cellId": "r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "qwen3.8-27b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm",
      "cell": {
        "id": "r1-btl4-35b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 1
      },
      "startedAt": "2026-10-20T00:52:55.513Z",
      "finishedAt": "2026-10-20T00:57:24.071Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 268558,
        "timing": {
          "startedAt": "2026-10-20T00:52:55.513Z",
          "finishedAt": "2026-10-20T00:57:24.071Z",
          "timeToFirstArtifactMs": 263496,
          "timeToLastArtifactWriteMs": 263500,
          "timeToFirstTokenMs": 244043,
          "firstTurnTtftMs": 8684,
          "timeToFirstToolCallMs": 246991
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 112.2,
        "prefillTokensPerSec": 2926.16,
        "decodeTokens": 1699,
        "peakRssMb": 27962,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:01:20.004Z",
        "trialId": "tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm",
        "cellId": "r1-btl4-35b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 3897,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "Clean, compact dark board with distinct X/O colors and large targets. Local browser verifies all eight wins, O win, draw, occupied-cell protection, terminal lock and restart; fits 375px. Readable static-button code, but empty cells lack accessible names and status has no live region. No winning-line highlight, scorekeeping or distinctive finishing touches.",
        "artifactSha256": "021170e6d5de159ad039f2e7efb227d9bc42e3deb2ddad77136883b026b45384",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv",
      "cell": {
        "id": "r1-btl4-35b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 1
      },
      "startedAt": "2026-10-20T00:57:59.050Z",
      "finishedAt": "2026-10-20T01:02:47.820Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 288770,
        "timing": {
          "startedAt": "2026-10-20T00:57:59.050Z",
          "finishedAt": "2026-10-20T01:02:47.820Z",
          "timeToFirstArtifactMs": 268969,
          "timeToLastArtifactWriteMs": 268973,
          "timeToFirstTokenMs": 244639,
          "firstTurnTtftMs": 8680,
          "timeToFirstToolCallMs": 252365
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 103.45,
        "prefillTokensPerSec": 2449.67,
        "decodeTokens": 2582,
        "peakRssMb": 37098,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:16:45.810Z",
        "trialId": "petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv",
        "cellId": "r1-btl4-35b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/workspace/pet-shop-website/index.html",
        "artifactBytes": 3881,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 2,
          "functionalCompleteness": 3,
          "codeQuality": 3,
          "polish": 1
        },
        "meanScore": 2.3,
        "justification": "HTML correctly preceded AI-logo generation in one workspace. Shipping styles.css and app.js are absent: browser shows unstyled text, two oversized distorted paw logos, empty pet/service sections and inert cart controls. At 375px the page overflows to 520px. Semantic scaffold and real PNG exist, but this is unfinished rather than a usable shop.",
        "artifactSha256": "fd6706aa28279e564380da790f043ffca82023c3c958006e6a0e34f2c638aeb7",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm",
      "cell": {
        "id": "r1-btl4-35b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:03:27.386Z",
      "finishedAt": "2026-10-20T01:08:56.617Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 329231,
        "timing": {
          "startedAt": "2026-10-20T01:03:27.386Z",
          "finishedAt": "2026-10-20T01:08:56.617Z",
          "timeToFirstArtifactMs": 292574,
          "timeToLastArtifactWriteMs": 320440,
          "timeToFirstTokenMs": 243694,
          "firstTurnTtftMs": 8796,
          "timeToFirstToolCallMs": 248341
        },
        "toolCalls": 6,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 101.01,
        "prefillTokensPerSec": 2431.67,
        "decodeTokens": 5369,
        "peakRssMb": 29769,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:16:45.811Z",
        "trialId": "tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm",
        "cellId": "r1-btl4-35b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 10956,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 5,
          "functionalCompleteness": 7,
          "codeQuality": 6,
          "polish": 5
        },
        "meanScore": 5.8,
        "justification": "Browser confirms movement/fire, natural enemy spawning, 100-point hits, enemy shots, lives, game over and restart. Grid, HUD and particles provide basic arcade feedback. Visible gun faces 90 degrees away from shot direction; movement and timers depend on frame rate. Fixed 806px canvas and HUD clip badly at 375px. Readable code, modest finish.",
        "artifactSha256": "66c4b90db0b24b3d201e3a04cc0b3191e2417fdd5567bd2a780ef56b11ea9efb",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh",
      "cell": {
        "id": "r1-btl4-35b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:09:32.151Z",
      "finishedAt": "2026-10-20T01:14:43.502Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 311351,
        "timing": {
          "startedAt": "2026-10-20T01:09:32.151Z",
          "finishedAt": "2026-10-20T01:14:43.502Z",
          "timeToFirstArtifactMs": 242684,
          "timeToLastArtifactWriteMs": 309447,
          "timeToFirstTokenMs": 241732,
          "firstTurnTtftMs": 7116,
          "timeToFirstToolCallMs": 242686
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 107.5,
        "prefillTokensPerSec": 2481.29,
        "decodeTokens": 3960,
        "peakRssMb": 30966,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:15:11.266Z",
        "trialId": "schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh",
        "cellId": "r1-btl4-35b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o",
      "cell": {
        "id": "r1-btl4-35b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:14:52.858Z",
      "finishedAt": "2026-10-20T01:23:15.943Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 503085,
        "timing": {
          "startedAt": "2026-10-20T01:14:52.858Z",
          "finishedAt": "2026-10-20T01:23:15.943Z",
          "timeToFirstArtifactMs": 249720,
          "timeToLastArtifactWriteMs": 499216,
          "timeToFirstTokenMs": 242602,
          "firstTurnTtftMs": 7605,
          "timeToFirstToolCallMs": 243259
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 115.37,
        "prefillTokensPerSec": 2902.38,
        "decodeTokens": 1113,
        "peakRssMb": 28292,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:30:10.786Z",
        "trialId": "failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o",
        "cellId": "r1-btl4-35b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo",
      "cell": {
        "id": "r1-btl4-35b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:23:24.971Z",
      "finishedAt": "2026-10-20T01:27:49.320Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 264349,
        "timing": {
          "startedAt": "2026-10-20T01:23:24.971Z",
          "finishedAt": "2026-10-20T01:27:49.320Z",
          "timeToFirstArtifactMs": 259986,
          "timeToLastArtifactWriteMs": 259986,
          "timeToFirstTokenMs": 246547,
          "firstTurnTtftMs": 7637,
          "timeToFirstToolCallMs": 247237
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 113.32,
        "prefillTokensPerSec": 2951.1,
        "decodeTokens": 1460,
        "peakRssMb": 27103,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:30:10.787Z",
        "trialId": "symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo",
        "cellId": "r1-btl4-35b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g",
      "cell": {
        "id": "r1-btl4-35b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:27:58.250Z",
      "finishedAt": "2026-10-20T01:33:19.654Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 321404,
        "timing": {
          "startedAt": "2026-10-20T01:27:58.250Z",
          "finishedAt": "2026-10-20T01:33:19.654Z",
          "timeToFirstArtifactMs": 264949,
          "timeToLastArtifactWriteMs": 310609,
          "timeToFirstTokenMs": 243005,
          "firstTurnTtftMs": 7717,
          "timeToFirstToolCallMs": 244248
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 102.07,
        "prefillTokensPerSec": 2189.21,
        "decodeTokens": 7167,
        "peakRssMb": 27892,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:45:19.123Z",
        "trialId": "data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g",
        "cellId": "r1-btl4-35b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k",
      "cell": {
        "id": "r1-btl4-35b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:33:29.511Z",
      "finishedAt": "2026-10-20T01:39:05.626Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 336115,
        "timing": {
          "startedAt": "2026-10-20T01:33:29.511Z",
          "finishedAt": "2026-10-20T01:39:05.626Z",
          "timeToFirstArtifactMs": 275027,
          "timeToLastArtifactWriteMs": 331566,
          "timeToFirstTokenMs": 245290,
          "firstTurnTtftMs": 10349,
          "timeToFirstToolCallMs": 246868
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 96.52,
        "prefillTokensPerSec": 2192.22,
        "decodeTokens": 7152,
        "peakRssMb": 28819,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:45:19.124Z",
        "trialId": "incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k",
        "cellId": "r1-btl4-35b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav",
      "cell": {
        "id": "r1-btl4-35b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:39:15.329Z",
      "finishedAt": "2026-10-20T01:43:41.118Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265789,
        "timing": {
          "startedAt": "2026-10-20T01:39:15.329Z",
          "finishedAt": "2026-10-20T01:43:41.118Z",
          "timeToFirstArtifactMs": 248854,
          "timeToLastArtifactWriteMs": 260304,
          "timeToFirstTokenMs": 245342,
          "firstTurnTtftMs": 10241,
          "timeToFirstToolCallMs": 246058
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 108.7,
        "prefillTokensPerSec": 2548.99,
        "decodeTokens": 1456,
        "peakRssMb": 27265,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T01:45:19.124Z",
        "trialId": "ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav",
        "cellId": "r1-btl4-35b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx",
      "cell": {
        "id": "r1-btl4-35b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:43:50.892Z",
      "finishedAt": "2026-10-20T01:48:31.615Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 280723,
        "timing": {
          "startedAt": "2026-10-20T01:43:50.892Z",
          "finishedAt": "2026-10-20T01:48:31.615Z",
          "timeToFirstArtifactMs": 253601,
          "timeToLastArtifactWriteMs": 276527,
          "timeToFirstTokenMs": 242656,
          "firstTurnTtftMs": 7756,
          "timeToFirstToolCallMs": 243310
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 113.14,
        "prefillTokensPerSec": 2533.46,
        "decodeTokens": 2258,
        "peakRssMb": 28769,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:00:11.929Z",
        "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx",
        "cellId": "r1-btl4-35b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk",
      "cell": {
        "id": "r1-btl4-35b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:48:41.400Z",
      "finishedAt": "2026-10-20T01:55:02.689Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/1f501e79 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model couldn't apply its file edit after 5 tries, so the turn was stopped. Try sending your message again, or switch to a more capable model in Settings.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 381289,
        "timing": {
          "startedAt": "2026-10-20T01:48:41.400Z",
          "finishedAt": "2026-10-20T01:55:02.689Z",
          "timeToFirstArtifactMs": 272064,
          "timeToLastArtifactWriteMs": 358563,
          "timeToFirstTokenMs": 245211,
          "firstTurnTtftMs": 10242,
          "timeToFirstToolCallMs": 246741
        },
        "toolCalls": 45,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.09,
        "prefillTokensPerSec": 2137.73,
        "decodeTokens": 9554,
        "peakRssMb": 33509,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:00:11.929Z",
        "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk",
        "cellId": "r1-btl4-35b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:55:12.671Z",
      "finishedAt": "2026-10-20T01:59:31.692Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 259021,
        "timing": {
          "startedAt": "2026-10-20T01:55:12.671Z",
          "finishedAt": "2026-10-20T01:59:31.692Z",
          "timeToFirstArtifactMs": 258065,
          "timeToLastArtifactWriteMs": 258065,
          "timeToFirstTokenMs": 245663,
          "firstTurnTtftMs": 7026,
          "timeToFirstToolCallMs": 248540
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 118.07,
        "prefillTokensPerSec": 2924.03,
        "decodeTokens": 1058,
        "peakRssMb": 28185,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:00:11.929Z",
        "trialId": "constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew",
        "cellId": "r1-btl4-35b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 1
      },
      "startedAt": "2026-10-20T01:59:40.749Z",
      "finishedAt": "2026-10-20T02:05:13.128Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 332379,
        "timing": {
          "startedAt": "2026-10-20T01:59:40.749Z",
          "finishedAt": "2026-10-20T02:05:13.128Z",
          "timeToFirstArtifactMs": 272487,
          "timeToLastArtifactWriteMs": 329217,
          "timeToFirstTokenMs": 243182,
          "firstTurnTtftMs": 7797,
          "timeToFirstToolCallMs": 244365
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 104.83,
        "prefillTokensPerSec": 2391.19,
        "decodeTokens": 7082,
        "peakRssMb": 29303,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:12:11.101Z",
        "trialId": "records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy",
        "cellId": "r1-btl4-35b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:05:24.210Z",
      "finishedAt": "2026-10-20T02:09:39.892Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 255682,
        "timing": {
          "startedAt": "2026-10-20T02:05:24.210Z",
          "finishedAt": "2026-10-20T02:09:39.892Z",
          "timeToFirstArtifactMs": 252304,
          "timeToLastArtifactWriteMs": 252304,
          "timeToFirstTokenMs": 243049,
          "firstTurnTtftMs": 7756,
          "timeToFirstToolCallMs": 243703
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 113.04,
        "prefillTokensPerSec": 2817.12,
        "decodeTokens": 1148,
        "peakRssMb": 27136,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:12:11.101Z",
        "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r",
        "cellId": "r1-btl4-35b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:09:48.993Z",
      "finishedAt": "2026-10-20T02:15:35.839Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 346846,
        "timing": {
          "startedAt": "2026-10-20T02:09:48.993Z",
          "finishedAt": "2026-10-20T02:15:35.839Z",
          "timeToFirstArtifactMs": 254485,
          "timeToLastArtifactWriteMs": 346030,
          "timeToFirstTokenMs": 245713,
          "firstTurnTtftMs": 9565,
          "timeToFirstToolCallMs": 246848
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 105.57,
        "prefillTokensPerSec": 2540.71,
        "decodeTokens": 1213,
        "peakRssMb": 27456,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:33:13.482Z",
        "trialId": "meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq",
        "cellId": "r1-btl4-35b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1492,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 9,
          "grounding": 9,
          "actionability": 9,
          "brevity": 9
        },
        "meanScore": 9,
        "justification": "Compact, clearly separated decisions, four executable actions, open questions and risks. Sep 14/EU-only/overnight migration, all roster owners and dates, A2→A1 dependency and exact CSV agree with the transcript. Jordan schedules the review by Aug 30 without implying the decision occurs then. Stale global/rolling/Casey proposals stay rejected; legal and rollback gaps remain explicit.",
        "artifactSha256": "17a9d9a6d22b8a4b9ef86a06be365a97e94094c2aaf200644177a1e4772287fb",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:15:48.887Z",
      "finishedAt": "2026-10-20T02:25:39.117Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 590230,
        "timing": {
          "startedAt": "2026-10-20T02:15:48.887Z",
          "finishedAt": "2026-10-20T02:25:39.117Z",
          "timeToFirstArtifactMs": 269252,
          "timeToLastArtifactWriteMs": 525518,
          "timeToFirstTokenMs": 249097,
          "firstTurnTtftMs": 8847,
          "timeToFirstToolCallMs": 251173
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 75.9,
        "prefillTokensPerSec": 1587.77,
        "decodeTokens": 23100,
        "peakRssMb": 30184,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:33:13.484Z",
        "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k",
        "cellId": "r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 3318,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 2
        },
        "meanScore": 4,
        "justification": "Scope preceded build and evaluate→finish, but verification falsely claims $1,375 matches runtime $20,980. Scope also miscalculates 60×120×85% as $8,160 instead of $6,120. CAC divides $25 by customers rather than marketing spend, inflating LTV:CAC to 18,178.6×. Missing totals produce NaN and a toFixed exception, truncating output rows. Editable zones cannot make this model reliable.",
        "artifactSha256": "713b91396489b8fcc6203c2e6a72354a026f4f666c4c74592de08548cc7b6757",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:25:49.044Z",
      "finishedAt": "2026-10-20T02:31:51.298Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/ee029de4 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model couldn't apply its file edit after 5 tries, so the turn was stopped. Try sending your message again, or switch to a more capable model in Settings.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 84.6% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 362254,
        "timing": {
          "startedAt": "2026-10-20T02:25:49.044Z",
          "finishedAt": "2026-10-20T02:31:51.298Z",
          "timeToFirstArtifactMs": 286821,
          "timeToLastArtifactWriteMs": 307968,
          "timeToFirstTokenMs": 245983,
          "firstTurnTtftMs": 10300,
          "timeToFirstToolCallMs": 247833
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 87.18,
        "prefillTokensPerSec": 1695.24,
        "decodeTokens": 6902,
        "peakRssMb": 31839,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:33:21.812Z",
        "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho",
        "cellId": "r1-btl4-35b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:32:18.424Z",
      "finishedAt": "2026-10-20T02:36:59.568Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 281144,
        "timing": {
          "startedAt": "2026-10-20T02:32:18.424Z",
          "finishedAt": "2026-10-20T02:36:59.568Z",
          "timeToFirstArtifactMs": 264640,
          "timeToLastArtifactWriteMs": 275337,
          "timeToFirstTokenMs": 244382,
          "firstTurnTtftMs": 8809,
          "timeToFirstToolCallMs": 244880
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 106.61,
        "prefillTokensPerSec": 2514.28,
        "decodeTokens": 2114,
        "peakRssMb": 30760,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:46:36.396Z",
        "trialId": "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro",
        "cellId": "r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1000,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 6,
          "candor": 9,
          "usefulness": 6
        },
        "meanScore": 7,
        "justification": "Theme ID, fonts, colors and 11→9 pages match tools; skipped Callout Emphasis and retained source formatting are explicit. Report invents one slide collapsing, sidebar usage and certainty that brand reviewers will flag it. Bold/accent remapping is actionable, but calling it the brand emphasis level needs approval. Tool evidence supports tighter spacing, not that detailed layout story.",
        "artifactSha256": "5c48c2254ea04d5046e9eae4fc67b8dccedc95b6a499a8c46d1332a79b53b11a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:37:09.884Z",
      "finishedAt": "2026-10-20T02:43:56.847Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 406963,
        "timing": {
          "startedAt": "2026-10-20T02:37:09.884Z",
          "finishedAt": "2026-10-20T02:43:56.847Z",
          "timeToFirstArtifactMs": 263792,
          "timeToLastArtifactWriteMs": 403251,
          "timeToFirstTokenMs": 250279,
          "firstTurnTtftMs": 10014,
          "timeToFirstToolCallMs": 251046
        },
        "toolCalls": 46,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 96.59,
        "prefillTokensPerSec": 2177.25,
        "decodeTokens": 10671,
        "peakRssMb": 33499,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T02:46:36.397Z",
        "trialId": "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3",
        "cellId": "r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 941,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 6,
          "convertibility": 6
        },
        "meanScore": 7.5,
        "justification": "Exact brief was read/cited before outlining; all seven H1 titles match in order. Metrics/actions survive, but warehouse exclusion, instant-email timing and floor-ops mismatch handling are unsourced. Cover says only audience; closing is empty, and [3] references are undefined. Conversion/preview/save/copy receipts show matching real PPTX bytes, but source content needs finishing.",
        "artifactSha256": "5938f0ae562a9056d8ee73701d6db312fae2afee26f12468e71e028b1f9061ad",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:44:06.603Z",
      "finishedAt": "2026-10-20T02:50:13.065Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: word-band. Last failure: research brief is 558 words; required range is 700-1,500",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 366462,
        "timing": {
          "startedAt": "2026-10-20T02:44:06.603Z",
          "finishedAt": "2026-10-20T02:50:13.065Z",
          "timeToFirstArtifactMs": 285713,
          "timeToLastArtifactWriteMs": 359153,
          "timeToFirstTokenMs": 242501,
          "firstTurnTtftMs": 7292,
          "timeToFirstToolCallMs": 249193
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 101.07,
        "prefillTokensPerSec": 2040.55,
        "decodeTokens": 7830,
        "peakRssMb": 30415,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T03:02:55.825Z",
        "trialId": "wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex",
        "cellId": "r1-btl4-35b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 4234,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 6,
          "citationQuality": 8,
          "calibration": 6
        },
        "meanScore": 6,
        "justification": "Readable but repetitive source summaries rather than a developed argument; below the 700-word brief. Dates, design and nearby source ids are faithful. Correctly disputes first-programmer shorthand, but infers the algorithm was never executed and unverified implementation caused the authorship debate, neither established by the signed cards.",
        "artifactSha256": "15ea39f3745ba7f2daedf3d8c8c2cd4a0f248e7abbfa5e7980f5986b080bc8b0",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 1
      },
      "startedAt": "2026-10-20T02:50:22.961Z",
      "finishedAt": "2026-10-20T03:24:41.411Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "craftbook-week-plan stale no-write repair loop: tasks/eval/report.md had unchanged content (3065 bytes) for 61s — a re-emission of identical bytes counts as unchanged here after repair feedback and newer chat activity; still failing 10/12: tasks/eval/report.md is missing required content: the agenda-less vendor sync from the live calendar is in the plan — nothing in its 3065 bytes matches /Vendor sync/. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 83.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2058450,
        "timing": {
          "startedAt": "2026-10-20T02:50:22.961Z",
          "finishedAt": "2026-10-20T03:24:41.411Z",
          "timeToFirstArtifactMs": 1013145,
          "timeToLastArtifactWriteMs": 2023478,
          "timeToFirstTokenMs": 252042,
          "firstTurnTtftMs": 11538,
          "timeToFirstToolCallMs": 253301
        },
        "toolCalls": 168,
        "autoAnswers": 8,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 59.17,
        "prefillTokensPerSec": 1188.25,
        "decodeTokens": 61428,
        "peakRssMb": 40231,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T03:30:50.364Z",
        "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8",
        "cellId": "r1-btl4-35b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/workspace/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 3107,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 0,
          "prioritization": 2,
          "conflict-handling": 0,
          "realism": 0
        },
        "meanScore": 0.5,
        "justification": "Drops all eight fetched events and falsely calls the calendar unreachable/empty despite two successful probes. Misses both prep meetings and the Tuesday overlap; 09–11 focus blocks collide with standup, budget/1:1 and design review. Follow-ups retain correct dates and finish before deadlines, but the claimed light week and committed blocks are unreliable.",
        "artifactSha256": "2925545e652c6f723fd15ea0cd4436008233a77263bd96c34d015419f61a756a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 1
      },
      "startedAt": "2026-10-20T03:24:56.390Z",
      "finishedAt": "2026-10-20T03:40:33.646Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 937256,
        "timing": {
          "startedAt": "2026-10-20T03:24:56.390Z",
          "finishedAt": "2026-10-20T03:40:33.646Z",
          "timeToFirstArtifactMs": 282553,
          "timeToLastArtifactWriteMs": 935729,
          "timeToFirstTokenMs": 251705,
          "firstTurnTtftMs": 11148,
          "timeToFirstToolCallMs": 252630
        },
        "toolCalls": 91,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 51.64,
        "prefillTokensPerSec": 1195.22,
        "decodeTokens": 26234,
        "peakRssMb": 37742,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T03:45:11.433Z",
        "trialId": "craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn",
        "cellId": "r1-btl4-35b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2365,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 1
      },
      "startedAt": "2026-10-20T03:40:51.772Z",
      "finishedAt": "2026-10-20T03:48:02.306Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 430534,
        "timing": {
          "startedAt": "2026-10-20T03:40:51.772Z",
          "finishedAt": "2026-10-20T03:48:02.306Z",
          "timeToFirstArtifactMs": 258825,
          "timeToLastArtifactWriteMs": 362211,
          "timeToFirstTokenMs": 250576,
          "firstTurnTtftMs": 9940,
          "timeToFirstToolCallMs": 251175
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 70.16,
        "prefillTokensPerSec": 1679.08,
        "decodeTokens": 9334,
        "peakRssMb": 32034,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:00:51.303Z",
        "trialId": "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj",
        "cellId": "r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 8722,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 9,
          "grounding": 3,
          "structure": 9,
          "actionability": 6
        },
        "meanScore": 6.8,
        "justification": "All six citations and numeric findings match, with consistent Summary/Evaluation/Relevance sections. Numerous unsupported details weaken the critique: single-site quality-improvement status despite four clinics, usual-care comparators, independence from motivation/device access, purposive sampling, self-reported adherence and alleged review gaps. Program implications are useful but mostly broad.",
        "artifactSha256": "234060458f70a51e26b12b27b59d8f286a9e62cdd9e8d2ddc4060c319c94effb",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk",
      "cell": {
        "id": "r1-btl4-35b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 1
      },
      "startedAt": "2026-10-20T03:48:17.438Z",
      "finishedAt": "2026-10-20T03:54:24.245Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 366807,
        "timing": {
          "startedAt": "2026-10-20T03:48:17.438Z",
          "finishedAt": "2026-10-20T03:54:24.245Z",
          "timeToFirstArtifactMs": 257645,
          "timeToLastArtifactWriteMs": 363887,
          "timeToFirstTokenMs": 250203,
          "firstTurnTtftMs": 9337,
          "timeToFirstToolCallMs": 250851
        },
        "toolCalls": 40,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.97,
        "prefillTokensPerSec": 2436.93,
        "decodeTokens": 6556,
        "peakRssMb": 33272,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:00:51.303Z",
        "trialId": "craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk",
        "cellId": "r1-btl4-35b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 2014,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 5,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 7,
        "justification": "Retains all five numeric values, the top complaint and three actions in complete conversion-ready Markdown. Adds unsupported headcount/customer-satisfaction effects, days-to-hours processing, causal success/scaling claims and an immediate photo-trigger email policy. Clear title, summary, metrics table and next actions; tool receipts show DOCX conversion, preview and save.",
        "artifactSha256": "4d4ff9c446e0947afe0fa822bda30691a28c32e02558063f7cedfe8fecf0e140",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 1
      },
      "startedAt": "2026-10-20T03:54:40.084Z",
      "finishedAt": "2026-10-20T03:59:16.325Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 276241,
        "timing": {
          "startedAt": "2026-10-20T03:54:40.084Z",
          "finishedAt": "2026-10-20T03:59:16.325Z",
          "timeToFirstArtifactMs": 273211,
          "timeToLastArtifactWriteMs": 273232,
          "timeToFirstTokenMs": 243012,
          "firstTurnTtftMs": 7582,
          "timeToFirstToolCallMs": 244076
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 105.75,
        "prefillTokensPerSec": 2850.95,
        "decodeTokens": 3083,
        "peakRssMb": 27117,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:00:12.320Z",
        "trialId": "fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu",
        "cellId": "r1-btl4-35b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 1
      },
      "startedAt": "2026-10-20T03:59:26.104Z",
      "finishedAt": "2026-10-20T04:03:41.843Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 255739,
        "timing": {
          "startedAt": "2026-10-20T03:59:26.104Z",
          "finishedAt": "2026-10-20T04:03:41.843Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 245849,
          "firstTurnTtftMs": 10565,
          "timeToFirstToolCallMs": 247688
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 107.81,
        "prefillTokensPerSec": 2640.39,
        "decodeTokens": 858,
        "peakRssMb": 27407,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:15:11.144Z",
        "trialId": "dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi",
        "cellId": "r1-btl4-35b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 1
      },
      "startedAt": "2026-10-20T04:03:51.004Z",
      "finishedAt": "2026-10-20T04:08:22.297Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 4 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 271293,
        "timing": {
          "startedAt": "2026-10-20T04:03:51.004Z",
          "finishedAt": "2026-10-20T04:08:22.297Z",
          "timeToFirstArtifactMs": 268557,
          "timeToLastArtifactWriteMs": 268581,
          "timeToFirstTokenMs": 244277,
          "firstTurnTtftMs": 8773,
          "timeToFirstToolCallMs": 249282
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 110.87,
        "prefillTokensPerSec": 2807.82,
        "decodeTokens": 1966,
        "peakRssMb": 28003,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:15:11.144Z",
        "trialId": "interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk",
        "cellId": "r1-btl4-35b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 1
      },
      "startedAt": "2026-10-20T04:08:32.569Z",
      "finishedAt": "2026-10-20T04:14:09.214Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 336645,
        "timing": {
          "startedAt": "2026-10-20T04:08:32.569Z",
          "finishedAt": "2026-10-20T04:14:09.214Z",
          "timeToFirstArtifactMs": 305033,
          "timeToLastArtifactWriteMs": 305033,
          "timeToFirstTokenMs": 249092,
          "firstTurnTtftMs": 8630,
          "timeToFirstToolCallMs": 249881
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 93.57,
        "prefillTokensPerSec": 2081.97,
        "decodeTokens": 4731,
        "peakRssMb": 29997,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:15:43.471Z",
        "trialId": "craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7",
        "cellId": "r1-btl4-35b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3015,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 8
        },
        "meanScore": 8,
        "justification": "Runs scope, gated artifact report and terminal verdict. Finds SQL injection, debug logging, handle leakage, pagination and false-success settings, with concrete locations/fixes and no findings against allowed patterns. Missing amount validation predates this change. Scope misdescribes commit mode as unsaved work; final note says seven findings although the report and severity counts total six.",
        "artifactSha256": "a07a07dc4ad1d5068ac33d71610b21e1977e1d58b5aa33f92750a77b34866bc3",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 1
      },
      "startedAt": "2026-10-20T04:14:21.587Z",
      "finishedAt": "2026-10-20T04:26:42.324Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 740737,
        "timing": {
          "startedAt": "2026-10-20T04:14:21.587Z",
          "finishedAt": "2026-10-20T04:26:42.324Z",
          "timeToFirstArtifactMs": 271124,
          "timeToLastArtifactWriteMs": 631720,
          "timeToFirstTokenMs": 249054,
          "firstTurnTtftMs": 9066,
          "timeToFirstToolCallMs": 250559
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 62.7,
        "prefillTokensPerSec": 1484.36,
        "decodeTokens": 27088,
        "peakRssMb": 29781,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:30:39.645Z",
        "trialId": "craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs",
        "cellId": "r1-btl4-35b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 16999,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 4
        },
        "meanScore": 6,
        "justification": "Completes the staged review with substantial corrected snippets and catches numeric retryCount drift. Misses the absent history handler and undocumented single-widget GET, while falsely treating optional nullable email omission as drift despite explicit conventions. Assumes 1.0.0 means no deployed clients; proposed references are incomplete, and duplicated issues/counts obscure priorities.",
        "artifactSha256": "89edc146538af7d734dcf6eae5b5e080eca72bca06d15c1c3e2dbcb551c3c9c5",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 1
      },
      "startedAt": "2026-10-20T04:26:54.580Z",
      "finishedAt": "2026-10-20T04:32:57.921Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 363341,
        "timing": {
          "startedAt": "2026-10-20T04:26:54.580Z",
          "finishedAt": "2026-10-20T04:32:57.921Z",
          "timeToFirstArtifactMs": 279052,
          "timeToLastArtifactWriteMs": 357900,
          "timeToFirstTokenMs": 245622,
          "firstTurnTtftMs": 10112,
          "timeToFirstToolCallMs": 247053
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 89.75,
        "prefillTokensPerSec": 1967.55,
        "decodeTokens": 7856,
        "peakRssMb": 30265,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T04:45:38.405Z",
        "trialId": "craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula",
        "cellId": "r1-btl4-35b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 5947,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 6,
          "remediation": 7
        },
        "meanScore": 7,
        "justification": "Confirms all three admin paths and clears documented safe constructs. Report/JSON severities and SQL/auth fixes agree. Overstates proven database writes, persistence and network reach. Filename filtering with a shell conflicts with the no-request-data-in-shell policy; the offered shell-free archive library is the stronger fix.",
        "artifactSha256": "5388a2d3ccb28de3942863ad0f83e2e99f1884144f012e59e2c18675e263a440",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 1
      },
      "startedAt": "2026-10-20T04:33:07.788Z",
      "finishedAt": "2026-10-20T04:52:47.566Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1179778,
        "timing": {
          "startedAt": "2026-10-20T04:33:07.788Z",
          "finishedAt": "2026-10-20T04:52:47.566Z",
          "timeToFirstArtifactMs": 823130,
          "timeToLastArtifactWriteMs": 1175650,
          "timeToFirstTokenMs": 251646,
          "firstTurnTtftMs": 12067,
          "timeToFirstToolCallMs": 253779
        },
        "toolCalls": 34,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 47.16,
        "prefillTokensPerSec": 719.3,
        "decodeTokens": 18268,
        "peakRssMb": 33157,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T05:00:14.760Z",
        "trialId": "large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8",
        "cellId": "r1-btl4-35b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 1
      },
      "startedAt": "2026-10-20T04:53:19.105Z",
      "finishedAt": "2026-10-20T05:02:57.874Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 578769,
        "timing": {
          "startedAt": "2026-10-20T04:53:19.105Z",
          "finishedAt": "2026-10-20T05:02:57.874Z",
          "timeToFirstArtifactMs": 281943,
          "timeToLastArtifactWriteMs": 575423,
          "timeToFirstTokenMs": 257047,
          "firstTurnTtftMs": 16053,
          "timeToFirstToolCallMs": 259105
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 58.28,
        "prefillTokensPerSec": 1055.16,
        "decodeTokens": 13107,
        "peakRssMb": 35379,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T05:15:41.467Z",
        "trialId": "craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb",
        "cellId": "r1-btl4-35b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 1659,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 7,
          "Invariant discipline": 9,
          "Reconciled verification": 6
        },
        "meanScore": 7.3,
        "justification": "Enumerates all seven sites from complete file reads before editing and preserves the changelog. All changes are mechanical and two real test runs pass 3/3. Counts reconcile, but the claimed exact residual re-scan has no successful search receipt: grep_artifact failed on workspace paths and the attempted grep package script was undefined. Manual rereads support the result, not the claimed scan.",
        "artifactSha256": "213bf266028828fa1696cf2e61db458ae22943de0e0f06db734a1a84c5cd4980",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 1
      },
      "startedAt": "2026-10-20T05:03:10.031Z",
      "finishedAt": "2026-10-20T05:41:00.707Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 725s — model busy but not delivering (hard digest stuck at c8a4ed880867f052)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 53.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2270676,
        "timing": {
          "startedAt": "2026-10-20T05:03:10.031Z",
          "finishedAt": "2026-10-20T05:41:00.707Z",
          "timeToFirstArtifactMs": 249953,
          "timeToLastArtifactWriteMs": 2168752,
          "timeToFirstTokenMs": 248624,
          "firstTurnTtftMs": 9422,
          "timeToFirstToolCallMs": 249957
        },
        "toolCalls": 54,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 55.95,
        "prefillTokensPerSec": 1289.77,
        "decodeTokens": 103843,
        "peakRssMb": 35290,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T05:45:11.955Z",
        "trialId": "craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls",
        "cellId": "r1-btl4-35b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "noArtifact",
        "comparisonEligible": false,
        "justification": "The exact declared artifact basename was absent from the captured workspace and artifacts drawer."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme",
      "cell": {
        "id": "r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 1
      },
      "startedAt": "2026-10-20T05:41:17.655Z",
      "finishedAt": "2026-10-20T05:49:00.569Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 462914,
        "timing": {
          "startedAt": "2026-10-20T05:41:17.655Z",
          "finishedAt": "2026-10-20T05:49:00.569Z",
          "timeToFirstArtifactMs": 253712,
          "timeToLastArtifactWriteMs": 458390,
          "timeToFirstTokenMs": 249794,
          "firstTurnTtftMs": 9350,
          "timeToFirstToolCallMs": 251027
        },
        "toolCalls": 63,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 88.15,
        "prefillTokensPerSec": 1901.72,
        "decodeTokens": 11827,
        "peakRssMb": 35200,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T06:01:05.361Z",
        "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme",
        "cellId": "r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1732,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 6,
          "Honest verification": 4
        },
        "meanScore": 6.3,
        "justification": "Real 80-vs-68 red run precedes the correct one-character >= fix, and a real green rerun follows. However, writing the reproduction overwrites all three original tests for small orders, large orders and invalid quantities. Fix notes omit the changed test file, claim nothing unverified despite an untested boundary, and the review accepts the reduced one-test suite as complete verification.",
        "artifactSha256": "a93486a29a50b90d00415fa46e7e5a9557c7a734e5a7f394444a7eb757b707f3",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 1
      },
      "startedAt": "2026-10-20T05:49:15.844Z",
      "finishedAt": "2026-10-20T05:53:34.356Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 258512,
        "timing": {
          "startedAt": "2026-10-20T05:49:15.844Z",
          "finishedAt": "2026-10-20T05:53:34.356Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 245705,
          "firstTurnTtftMs": 10616,
          "timeToFirstToolCallMs": 247363
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 108.49,
        "prefillTokensPerSec": 2642.45,
        "decodeTokens": 703,
        "peakRssMb": 27425,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T06:00:18.717Z",
        "trialId": "craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 1
      },
      "startedAt": "2026-10-20T05:53:43.474Z",
      "finishedAt": "2026-10-20T06:44:08.464Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on task quarterly-brief/1 craftbook; missing signals stayed unchanged: task quarterly-brief/1 still carries the original failing step prompt verbatim — read the task's craftbook, find what the final gate demands, and rewrite that step's prompt to instruct it, task quarterly-brief/1's craftbook still has 2 steps — add a verification step at the end of the recipe. Last failure: task quarterly-brief/1 still carries the original failing step prompt verbatim — read the task's craftbook, find what the final gate demands, and rewrite that step's prompt to instruct it",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3024990,
        "timing": {
          "startedAt": "2026-10-20T05:53:43.474Z",
          "finishedAt": "2026-10-20T06:44:08.464Z",
          "timeToFirstArtifactMs": 1531604,
          "timeToLastArtifactWriteMs": 3020605,
          "timeToFirstTokenMs": 248349,
          "firstTurnTtftMs": 13014,
          "timeToFirstToolCallMs": 248916
        },
        "toolCalls": 102,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 58.15,
        "prefillTokensPerSec": 1232.13,
        "decodeTokens": 132449,
        "peakRssMb": 36598,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:04:10.284Z",
        "trialId": "craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 1
      },
      "startedAt": "2026-10-20T06:44:19.926Z",
      "finishedAt": "2026-10-20T06:52:31.625Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 491699,
        "timing": {
          "startedAt": "2026-10-20T06:44:19.926Z",
          "finishedAt": "2026-10-20T06:52:31.625Z",
          "timeToFirstArtifactMs": 260376,
          "timeToLastArtifactWriteMs": 484445,
          "timeToFirstTokenMs": 249147,
          "firstTurnTtftMs": 9048,
          "timeToFirstToolCallMs": 249815
        },
        "toolCalls": 44,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 89.69,
        "prefillTokensPerSec": 2068.64,
        "decodeTokens": 15663,
        "peakRssMb": 33666,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:09:26.928Z",
        "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 596,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 7,
          "numbering continuity": 9,
          "professionalism": 2,
          "completeness": 4,
          "print-polish": 6
        },
        "meanScore": 5.6,
        "justification": "Ledger totals and 042–044 sequence are correct; all three render on one page with due dates matching terms. Invents addresses/contacts, zero tax, mixed dollar/euro currency, bank details and late interest; invoice 043 directs payment to the client. Books still say 041/uninvoiced despite the sent claim. Two print layouts retain a Print Invoice button; invoice styles differ.",
        "artifactSha256": "1959fb1d86a942766ad939fe587922c835cf6ff037e1aee1ddb9086ffeec2a83",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 1
      },
      "startedAt": "2026-10-20T06:52:40.699Z",
      "finishedAt": "2026-10-20T07:16:10.817Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on craftbook: CSV Order Cleanup; missing signals stayed unchanged: task order-intake-cleanup/1 (from craftbook \"csv-order-cleanup\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.. Last failure: task order-intake-cleanup/1 (from craftbook \"csv-order-cleanup\") has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1410118,
        "timing": {
          "startedAt": "2026-10-20T06:52:40.699Z",
          "finishedAt": "2026-10-20T07:16:10.817Z",
          "timeToFirstArtifactMs": 701861,
          "timeToLastArtifactWriteMs": 1407777,
          "timeToFirstTokenMs": 252852,
          "firstTurnTtftMs": 13140,
          "timeToFirstToolCallMs": 253931
        },
        "toolCalls": 57,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 55.97,
        "prefillTokensPerSec": 1147.6,
        "decodeTokens": 54352,
        "peakRssMb": 33180,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:04:10.285Z",
        "trialId": "craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 1
      },
      "startedAt": "2026-10-20T07:16:24.147Z",
      "finishedAt": "2026-10-20T07:31:35.778Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 911631,
        "timing": {
          "startedAt": "2026-10-20T07:16:24.147Z",
          "finishedAt": "2026-10-20T07:31:35.778Z",
          "timeToFirstArtifactMs": 328457,
          "timeToLastArtifactWriteMs": 635391,
          "timeToFirstTokenMs": 246285,
          "firstTurnTtftMs": 10591,
          "timeToFirstToolCallMs": 248401
        },
        "toolCalls": 56,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 73.86,
        "prefillTokensPerSec": 1439.48,
        "decodeTokens": 17326,
        "peakRssMb": 36278,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:04:10.285Z",
        "trialId": "craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 1
      },
      "startedAt": "2026-10-20T07:31:44.901Z",
      "finishedAt": "2026-10-20T09:18:23.012Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "did the Q1 pass, generalized it into a saved reusable craftbook, ran that craftbook to completion, and the Q2 outputs are genuinely derived from the Q2 list",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7.6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 7.5,
            "ruleId": "B_TWO_INTERVENTIONS",
            "summary": "No red flags or missing roles were observed; two interventions receive the midpoint between the fixed low/high-intervention anchors."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 6398111,
        "timing": {
          "startedAt": "2026-10-20T07:31:44.901Z",
          "finishedAt": "2026-10-20T09:18:23.012Z",
          "timeToFirstArtifactMs": 321361,
          "timeToLastArtifactWriteMs": 6395982,
          "timeToFirstTokenMs": 251859,
          "firstTurnTtftMs": 13180,
          "timeToFirstToolCallMs": 254593
        },
        "toolCalls": 270,
        "autoAnswers": 2,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 66.48,
        "prefillTokensPerSec": 1190.96,
        "decodeTokens": 215417,
        "peakRssMb": 38411,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:04:10.285Z",
        "trialId": "craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 1
      },
      "startedAt": "2026-10-20T09:18:34.283Z",
      "finishedAt": "2026-10-20T09:50:02.508Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-params:5:targetnone:fr1yhq739:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 42 tool calls (10 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.1,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 71.4% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1888225,
        "timing": {
          "startedAt": "2026-10-20T09:18:34.283Z",
          "finishedAt": "2026-10-20T09:50:02.508Z",
          "timeToFirstArtifactMs": 396081,
          "timeToLastArtifactWriteMs": 1654635,
          "timeToFirstTokenMs": 248728,
          "firstTurnTtftMs": 13091,
          "timeToFirstToolCallMs": 250744
        },
        "toolCalls": 80,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 66.63,
        "prefillTokensPerSec": 1247.05,
        "decodeTokens": 35605,
        "peakRssMb": 33189,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:04:10.285Z",
        "trialId": "craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 1
      },
      "startedAt": "2026-10-20T09:50:11.980Z",
      "finishedAt": "2026-10-20T10:22:25.139Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 3 while the failing detail kept changing; latest checked file craftbook: inventory report gate script. Last failure: gate script never calls gateResult(...) — the gate contract requires stamping a structured result",
        "failureMode": "model-stuck",
        "failureClass": "infra",
        "includedInFixedModelAggregate": false,
        "failureClassRule": "gate-unevaluable"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 60% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1933159,
        "timing": {
          "startedAt": "2026-10-20T09:50:11.980Z",
          "finishedAt": "2026-10-20T10:22:25.139Z",
          "timeToFirstArtifactMs": 854458,
          "timeToLastArtifactWriteMs": 1673453,
          "timeToFirstTokenMs": 248362,
          "firstTurnTtftMs": 13138,
          "timeToFirstToolCallMs": 251953
        },
        "toolCalls": 110,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 55.51,
        "prefillTokensPerSec": 1085.62,
        "decodeTokens": 57073,
        "peakRssMb": 36671,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T13:04:10.285Z",
        "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x",
      "cell": {
        "id": "r1-btl4-35b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 1
      },
      "startedAt": "2026-10-20T13:08:17.126Z",
      "finishedAt": "2026-10-20T14:00:30.879Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:7:targetnone:fr1tjwffw:rp0:rf0:m4\" stuck for 8m — 8 re-writes (34 tool calls) without sniff movement. Artifact produced but never reached success (retry-loop nudge was sent and ignored).",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.9,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 87.5% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3133753,
        "timing": {
          "startedAt": "2026-10-20T13:08:17.126Z",
          "finishedAt": "2026-10-20T14:00:30.879Z",
          "timeToFirstArtifactMs": 615999,
          "timeToLastArtifactWriteMs": 3024380,
          "timeToFirstTokenMs": 249741,
          "firstTurnTtftMs": 13290,
          "timeToFirstToolCallMs": 250579
        },
        "toolCalls": 198,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 67.43,
        "prefillTokensPerSec": 1351.68,
        "decodeTokens": 127173,
        "peakRssMb": 37691,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:11:50.295Z",
        "trialId": "craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x",
        "cellId": "r1-btl4-35b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt",
      "cell": {
        "id": "r2-btl4-35b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:00:43.799Z",
      "finishedAt": "2026-10-20T14:05:12.568Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 268769,
        "timing": {
          "startedAt": "2026-10-20T14:00:43.799Z",
          "finishedAt": "2026-10-20T14:05:12.568Z",
          "timeToFirstArtifactMs": 264279,
          "timeToLastArtifactWriteMs": 264283,
          "timeToFirstTokenMs": 244296,
          "firstTurnTtftMs": 8687,
          "timeToFirstToolCallMs": 248493
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 112.16,
        "prefillTokensPerSec": 2857.05,
        "decodeTokens": 1896,
        "peakRssMb": 27856,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:13:44.599Z",
        "trialId": "tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt",
        "cellId": "r2-btl4-35b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 4041,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 6
        },
        "meanScore": 6.8,
        "justification": "All eight win lines, O win, draw, occupied-cell guard, end lock and restart work without errors. A clear 312px board fits mobile; winning cells pulse and colors distinguish players. Simple readable state logic, but empty cells lack accessible names, status is not announced, and animation has no reduced-motion option.",
        "artifactSha256": "4548e3e63c5b74a7681e11a406a195f677cb961d9501e65e1ee96a42326ee2f6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh",
      "cell": {
        "id": "r2-btl4-35b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:05:47.784Z",
      "finishedAt": "2026-10-20T14:11:17.315Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 329531,
        "timing": {
          "startedAt": "2026-10-20T14:05:47.784Z",
          "finishedAt": "2026-10-20T14:11:17.315Z",
          "timeToFirstArtifactMs": 304116,
          "timeToLastArtifactWriteMs": 324654,
          "timeToFirstTokenMs": 244381,
          "firstTurnTtftMs": 8690,
          "timeToFirstToolCallMs": 249881
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 104.1,
        "prefillTokensPerSec": 2147.72,
        "decodeTokens": 5858,
        "peakRssMb": 37576,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:13:44.601Z",
        "trialId": "petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh",
        "cellId": "r2-btl4-35b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/workspace/pet-shop-website/index.html",
        "artifactBytes": 14239,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 7,
          "functionalCompleteness": 8,
          "codeQuality": 6,
          "polish": 6
        },
        "meanScore": 6.8,
        "justification": "Cohesive responsive shop with six products, four services and a real local AI logo; HTML correctly precedes image generation in one workspace. Cart count/toast and mobile menu work without errors or overflow. Generic emoji cards and placeholder social links limit finish; clickable product divs lack keyboard access, menu expanded state is absent, and aquarium kit is tagged Food.",
        "artifactSha256": "f0fccc9254899b9e699bd5b370aca143f82981b395b5a5b4989c64c2e909a9af",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp",
      "cell": {
        "id": "r2-btl4-35b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:11:58.422Z",
      "finishedAt": "2026-10-20T14:17:22.184Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 323762,
        "timing": {
          "startedAt": "2026-10-20T14:11:58.422Z",
          "finishedAt": "2026-10-20T14:17:22.184Z",
          "timeToFirstArtifactMs": 316694,
          "timeToLastArtifactWriteMs": 316701,
          "timeToFirstTokenMs": 243908,
          "firstTurnTtftMs": 8789,
          "timeToFirstToolCallMs": 249346
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.87,
        "prefillTokensPerSec": 2916.55,
        "decodeTokens": 6660,
        "peakRssMb": 27965,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:31:50.494Z",
        "trialId": "tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp",
        "cellId": "r2-btl4-35b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 17224,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 6,
          "polish": 6
        },
        "meanScore": 6.5,
        "justification": "Movement, firing, enemy hits, score, lives, pause and restart work; particles and rising difficulty add interest. The clean arcade grid is basic. Enemies pass through cover that blocks the player, all timing depends on frame rate, and the fixed canvas clips on phones.",
        "artifactSha256": "9fa183805c18064147af0643800e4605342d899f225cff64b6f1b1e8503223e8",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu",
      "cell": {
        "id": "r2-btl4-35b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:17:58.477Z",
      "finishedAt": "2026-10-20T14:22:59.246Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on src/migrate.ts; missing signals stayed unchanged: migrate-function, tests-present, migration-doc, tsc-clean. Last failure: migrate.ts harness-owned preservation gate failed at typecheck: <tmp>/gezel-eval-schema-migration-function-znH0Ml/src/migrate.ts(7,5): error TS2322: Type 'string' is not assignable to type 'number'.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.1,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 4.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 42.9% of its measured signals."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 300769,
        "timing": {
          "startedAt": "2026-10-20T14:17:58.477Z",
          "finishedAt": "2026-10-20T14:22:59.246Z",
          "timeToFirstArtifactMs": 247317,
          "timeToLastArtifactWriteMs": 299580,
          "timeToFirstTokenMs": 246308,
          "firstTurnTtftMs": 7111,
          "timeToFirstToolCallMs": 247319
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 112.1,
        "prefillTokensPerSec": 2559.44,
        "decodeTokens": 1253,
        "peakRssMb": 33294,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:26:46.901Z",
        "trialId": "schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu",
        "cellId": "r2-btl4-35b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd",
      "cell": {
        "id": "r2-btl4-35b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:23:13.871Z",
      "finishedAt": "2026-10-20T14:27:57.579Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 283708,
        "timing": {
          "startedAt": "2026-10-20T14:23:13.871Z",
          "finishedAt": "2026-10-20T14:27:57.579Z",
          "timeToFirstArtifactMs": 253766,
          "timeToLastArtifactWriteMs": 280893,
          "timeToFirstTokenMs": 244126,
          "firstTurnTtftMs": 8146,
          "timeToFirstToolCallMs": 244841
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.57,
        "prefillTokensPerSec": 2212.65,
        "decodeTokens": 1058,
        "peakRssMb": 29778,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:30:23.151Z",
        "trialId": "failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd",
        "cellId": "r2-btl4-35b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh",
      "cell": {
        "id": "r2-btl4-35b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:28:12.069Z",
      "finishedAt": "2026-10-20T14:32:57.617Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 285548,
        "timing": {
          "startedAt": "2026-10-20T14:28:12.069Z",
          "finishedAt": "2026-10-20T14:32:57.617Z",
          "timeToFirstArtifactMs": 257235,
          "timeToLastArtifactWriteMs": 282310,
          "timeToFirstTokenMs": 242552,
          "firstTurnTtftMs": 7635,
          "timeToFirstToolCallMs": 243229
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 106.31,
        "prefillTokensPerSec": 2847.59,
        "decodeTokens": 4151,
        "peakRssMb": 27106,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:41:43.696Z",
        "trialId": "symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh",
        "cellId": "r2-btl4-35b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9",
      "cell": {
        "id": "r2-btl4-35b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:33:07.793Z",
      "finishedAt": "2026-10-20T14:37:48.551Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 280758,
        "timing": {
          "startedAt": "2026-10-20T14:33:07.793Z",
          "finishedAt": "2026-10-20T14:37:48.551Z",
          "timeToFirstArtifactMs": 275129,
          "timeToLastArtifactWriteMs": 279192,
          "timeToFirstTokenMs": 242865,
          "firstTurnTtftMs": 7724,
          "timeToFirstToolCallMs": 243798
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 106.76,
        "prefillTokensPerSec": 2565.92,
        "decodeTokens": 3651,
        "peakRssMb": 27297,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:41:43.696Z",
        "trialId": "data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9",
        "cellId": "r2-btl4-35b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg",
      "cell": {
        "id": "r2-btl4-35b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:37:58.579Z",
      "finishedAt": "2026-10-20T14:46:01.823Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 483244,
        "timing": {
          "startedAt": "2026-10-20T14:37:58.579Z",
          "finishedAt": "2026-10-20T14:46:01.823Z",
          "timeToFirstArtifactMs": 480343,
          "timeToLastArtifactWriteMs": 480343,
          "timeToFirstTokenMs": 246029,
          "firstTurnTtftMs": 10395,
          "timeToFirstToolCallMs": 248236
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 80.35,
        "prefillTokensPerSec": 1903.47,
        "decodeTokens": 16272,
        "peakRssMb": 29575,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:56:45.608Z",
        "trialId": "incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg",
        "cellId": "r2-btl4-35b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc",
      "cell": {
        "id": "r2-btl4-35b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:46:12.996Z",
      "finishedAt": "2026-10-20T14:51:48.944Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on halt-report.md; the same scenario failure stayed unchanged. Last failure: halt-report.md must state the expected age threshold from runbook.md",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.9,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 85.7% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 335948,
        "timing": {
          "startedAt": "2026-10-20T14:46:12.996Z",
          "finishedAt": "2026-10-20T14:51:48.944Z",
          "timeToFirstArtifactMs": 249177,
          "timeToLastArtifactWriteMs": 334189,
          "timeToFirstTokenMs": 246031,
          "firstTurnTtftMs": 10908,
          "timeToFirstToolCallMs": 246605
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 99.75,
        "prefillTokensPerSec": 2155.03,
        "decodeTokens": 6558,
        "peakRssMb": 31624,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:56:45.609Z",
        "trialId": "ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc",
        "cellId": "r2-btl4-35b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm",
      "cell": {
        "id": "r2-btl4-35b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:52:03.408Z",
      "finishedAt": "2026-10-20T14:56:38.164Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 274756,
        "timing": {
          "startedAt": "2026-10-20T14:52:03.408Z",
          "finishedAt": "2026-10-20T14:56:38.164Z",
          "timeToFirstArtifactMs": 271754,
          "timeToLastArtifactWriteMs": 271754,
          "timeToFirstTokenMs": 250961,
          "firstTurnTtftMs": 12619,
          "timeToFirstToolCallMs": 252055
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 54.64,
        "prefillTokensPerSec": 1115.43,
        "decodeTokens": 1108,
        "peakRssMb": 27068,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T14:56:58.810Z",
        "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm",
        "cellId": "r2-btl4-35b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k",
      "cell": {
        "id": "r2-btl4-35b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-20T14:56:55.577Z",
      "finishedAt": "2026-10-20T15:02:47.572Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 351995,
        "timing": {
          "startedAt": "2026-10-20T14:56:55.577Z",
          "finishedAt": "2026-10-20T15:02:47.572Z",
          "timeToFirstArtifactMs": 275389,
          "timeToLastArtifactWriteMs": 346172,
          "timeToFirstTokenMs": 247030,
          "firstTurnTtftMs": 11097,
          "timeToFirstToolCallMs": 248575
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 95.17,
        "prefillTokensPerSec": 2045.2,
        "decodeTokens": 7092,
        "peakRssMb": 30830,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:11:43.918Z",
        "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k",
        "cellId": "r2-btl4-35b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:03:01.618Z",
      "finishedAt": "2026-10-20T15:07:47.624Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on customer-notice.md; the same scenario failure stayed unchanged. Last failure: word band violated: 132 words, need ≥ 140 — the notice must be 140-220 words",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_PARTIAL_UNKNOWN_MAX",
            "summary": "An authored, scored output exists, but the gate denominator is unavailable; partial credit is capped at the rubric midpoint."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 286006,
        "timing": {
          "startedAt": "2026-10-20T15:03:01.618Z",
          "finishedAt": "2026-10-20T15:07:47.624Z",
          "timeToFirstArtifactMs": 255181,
          "timeToLastArtifactWriteMs": 280255,
          "timeToFirstTokenMs": 242445,
          "firstTurnTtftMs": 7027,
          "timeToFirstToolCallMs": 246147
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 109.87,
        "prefillTokensPerSec": 2501.63,
        "decodeTokens": 2052,
        "peakRssMb": 30118,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:11:43.918Z",
        "trialId": "constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0",
        "cellId": "r2-btl4-35b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:08:01.333Z",
      "finishedAt": "2026-10-20T15:14:17.876Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 376543,
        "timing": {
          "startedAt": "2026-10-20T15:08:01.333Z",
          "finishedAt": "2026-10-20T15:14:17.876Z",
          "timeToFirstArtifactMs": 269929,
          "timeToLastArtifactWriteMs": 373146,
          "timeToFirstTokenMs": 242696,
          "firstTurnTtftMs": 7798,
          "timeToFirstToolCallMs": 243923
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 92,
        "prefillTokensPerSec": 2029.32,
        "decodeTokens": 11117,
        "peakRssMb": 28551,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:26:42.886Z",
        "trialId": "records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5",
        "cellId": "r2-btl4-35b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:14:30.934Z",
      "finishedAt": "2026-10-20T15:18:51.933Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 260999,
        "timing": {
          "startedAt": "2026-10-20T15:14:30.934Z",
          "finishedAt": "2026-10-20T15:18:51.933Z",
          "timeToFirstArtifactMs": 258102,
          "timeToLastArtifactWriteMs": 258102,
          "timeToFirstTokenMs": 243410,
          "firstTurnTtftMs": 7766,
          "timeToFirstToolCallMs": 244335
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 110.26,
        "prefillTokensPerSec": 2890.47,
        "decodeTokens": 1596,
        "peakRssMb": 27044,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:26:42.887Z",
        "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl",
        "cellId": "r2-btl4-35b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:19:04.406Z",
      "finishedAt": "2026-10-20T15:24:50.976Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 346570,
        "timing": {
          "startedAt": "2026-10-20T15:19:04.406Z",
          "finishedAt": "2026-10-20T15:24:50.976Z",
          "timeToFirstArtifactMs": 251815,
          "timeToLastArtifactWriteMs": 343657,
          "timeToFirstTokenMs": 244368,
          "firstTurnTtftMs": 8825,
          "timeToFirstToolCallMs": 245382
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 116.24,
        "prefillTokensPerSec": 2769.43,
        "decodeTokens": 1069,
        "peakRssMb": 27580,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:27:16.512Z",
        "trialId": "meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t",
        "cellId": "r2-btl4-35b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1425,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 9,
          "grounding": 9,
          "actionability": 9,
          "brevity": 9
        },
        "meanScore": 9,
        "justification": "Cleanly separates the September 14 EU-only overnight-batch decision, four dated actions, unresolved annual-plan and rollback ownership, and legal risk. Owners and A2→A1 dependency match the transcript and CSV; scheduling the annual-plan review is correctly distinguished from holding it. Compact and executable.",
        "artifactSha256": "00b104995cfd7008325e73b2484c7c635442fa239a830f35e33b07b8fda7f31b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:25:02.898Z",
      "finishedAt": "2026-10-20T15:31:41.961Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 399063,
        "timing": {
          "startedAt": "2026-10-20T15:25:02.898Z",
          "finishedAt": "2026-10-20T15:31:41.961Z",
          "timeToFirstArtifactMs": 282643,
          "timeToLastArtifactWriteMs": 350659,
          "timeToFirstTokenMs": 248845,
          "firstTurnTtftMs": 8831,
          "timeToFirstToolCallMs": 250712
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 94,
        "prefillTokensPerSec": 2315.32,
        "decodeTokens": 10659,
        "peakRssMb": 29451,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:43:17.081Z",
        "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w",
        "cellId": "r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 2448,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 4
        },
        "meanScore": 5.5,
        "justification": "Scopes before build and reaches review/finish, with live formulas. Yet the checklist calls $212 net while the hand-check gives -$1,288; recurring monthly opex is mislabeled as a 7.1-month payback despite continuing losses. Defaults also fail their own sanity indicator, and $100 AOV/$1,500 opex lack seeded support. Review misses these contradictions.",
        "artifactSha256": "d3e3e1e9ada4c254ffb2674684c31d1927c9d3ed28f8f691b269b34c429d1e52",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:32:00.462Z",
      "finishedAt": "2026-10-20T15:38:46.498Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 406036,
        "timing": {
          "startedAt": "2026-10-20T15:32:00.462Z",
          "finishedAt": "2026-10-20T15:38:46.498Z",
          "timeToFirstArtifactMs": 298995,
          "timeToLastArtifactWriteMs": 401446,
          "timeToFirstTokenMs": 249642,
          "firstTurnTtftMs": 10263,
          "timeToFirstToolCallMs": 251050
        },
        "toolCalls": 49,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 87.36,
        "prefillTokensPerSec": 1982.43,
        "decodeTokens": 9139,
        "peakRssMb": 36283,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:41:45.257Z",
        "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd",
        "cellId": "r2-btl4-35b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:38:58.541Z",
      "finishedAt": "2026-10-20T15:43:20.127Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 261586,
        "timing": {
          "startedAt": "2026-10-20T15:38:58.541Z",
          "finishedAt": "2026-10-20T15:43:20.127Z",
          "timeToFirstArtifactMs": 258956,
          "timeToLastArtifactWriteMs": 258956,
          "timeToFirstTokenMs": 244503,
          "firstTurnTtftMs": 8375,
          "timeToFirstToolCallMs": 244701
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 112,
        "prefillTokensPerSec": 2523.92,
        "decodeTokens": 1312,
        "peakRssMb": 28839,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:44:31.987Z",
        "trialId": "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa",
        "cellId": "r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1049,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 8,
          "candor": 9,
          "usefulness": 8
        },
        "meanScore": 8.3,
        "justification": "Theme values, page counts, five applied styles and preserved content match tool receipts. Callout Emphasis is plainly left unresolved, with a practical typography recommendation. The report overreaches by describing a specific page absorbing slide content; tools establish only tighter spacing and an 11-to-9 page reduction.",
        "artifactSha256": "6c78e706478e9d2b054b48d173c4f349895614787eebde3883b0c5ad1e338fe6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:43:29.025Z",
      "finishedAt": "2026-10-20T15:49:20.457Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 351432,
        "timing": {
          "startedAt": "2026-10-20T15:43:29.025Z",
          "finishedAt": "2026-10-20T15:49:20.457Z",
          "timeToFirstArtifactMs": 266243,
          "timeToLastArtifactWriteMs": 345647,
          "timeToFirstTokenMs": 250026,
          "firstTurnTtftMs": 9991,
          "timeToFirstToolCallMs": 250468
        },
        "toolCalls": 34,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 104.37,
        "prefillTokensPerSec": 2281.58,
        "decodeTokens": 6739,
        "peakRssMb": 31792,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:57:49.136Z",
        "trialId": "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz",
        "cellId": "r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1108,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 7,
          "convertibility": 9
        },
        "meanScore": 8.5,
        "justification": "Source read and cited before outlining; all six H1s match the outline, metrics are exact, and concise content converts to retained PPTX. Single-thread pilot scope, instant-photo email timing and the outline’s causal claim exceed the brief. A Sources slide follows Next Actions, weakening the requested action close.",
        "artifactSha256": "5c69ae805909916ec75bd363471d6ee973f561219946e1526ddb885b77ca74fa",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:49:30.365Z",
      "finishedAt": "2026-10-20T15:58:00.913Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on research-brief.md; missing signals stayed unchanged: word-band. Last failure: research brief is 658 words; required range is 700-1,500",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 510548,
        "timing": {
          "startedAt": "2026-10-20T15:49:30.365Z",
          "finishedAt": "2026-10-20T15:58:00.913Z",
          "timeToFirstArtifactMs": 369332,
          "timeToLastArtifactWriteMs": 507293,
          "timeToFirstTokenMs": 245920,
          "firstTurnTtftMs": 7296,
          "timeToFirstToolCallMs": 254057
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 89.48,
        "prefillTokensPerSec": 1827.21,
        "decodeTokens": 19197,
        "peakRssMb": 32490,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T15:58:51.512Z",
        "trialId": "wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe",
        "cellId": "r2-btl4-35b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5140,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 3,
          "citationQuality": 6,
          "calibration": 4
        },
        "meanScore": 4.3,
        "justification": "Correct chronology and key cards are cited, but mostly source-by-source repetition. Funding withdrawal, Babbage’s later priorities, never-executed claims and terminology assertions are absent from the cards; “first published computational procedure” overstates the evidence. The first-programmer caveat does not calibrate those additions. Final 658 words also miss the requested length.",
        "artifactSha256": "9fed98626639ac5707fa32dc626086f133ac322c33e17a2fb4ed8590225819c4",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 2
      },
      "startedAt": "2026-10-20T15:58:10.699Z",
      "finishedAt": "2026-10-20T17:32:09.867Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-week-plan passed 12 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 5639168,
        "timing": {
          "startedAt": "2026-10-20T15:58:10.699Z",
          "finishedAt": "2026-10-20T17:32:09.867Z",
          "timeToFirstArtifactMs": 1013871,
          "timeToLastArtifactWriteMs": 5635547,
          "timeToFirstTokenMs": 251460,
          "firstTurnTtftMs": 11437,
          "timeToFirstToolCallMs": 253050
        },
        "toolCalls": 899,
        "autoAnswers": 256,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 71.26,
        "prefillTokensPerSec": 1331.53,
        "decodeTokens": 209123,
        "peakRssMb": 41753,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T17:42:51.416Z",
        "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv",
        "cellId": "r2-btl4-35b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/workspace/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 3738,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 8,
          "prioritization": 5,
          "conflict-handling": 4,
          "realism": 5
        },
        "meanScore": 5.5,
        "justification": "All eight July events and three follow-ups survive a real calendar probe. Focus windows fit gaps but lack durations or allocation to the largest work. Neither agenda-less meeting is flagged for prep. Tuesday’s overlap is named, yet “pick one” gives no concrete reschedule; calling Tuesday busiest ignores Friday’s heavier meeting load.",
        "artifactSha256": "b3b00b7831854f45d941094a8af51f7bbbf38d224d0b63e190e246a314cff528",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 2
      },
      "startedAt": "2026-10-20T17:32:28.869Z",
      "finishedAt": "2026-10-20T17:41:40.846Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 551977,
        "timing": {
          "startedAt": "2026-10-20T17:32:28.869Z",
          "finishedAt": "2026-10-20T17:41:40.846Z",
          "timeToFirstArtifactMs": 262214,
          "timeToLastArtifactWriteMs": 550197,
          "timeToFirstTokenMs": 250480,
          "firstTurnTtftMs": 10633,
          "timeToFirstToolCallMs": 251671
        },
        "toolCalls": 41,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 96.51,
        "prefillTokensPerSec": 2361.38,
        "decodeTokens": 14293,
        "peakRssMb": 35234,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T17:42:51.622Z",
        "trialId": "craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm",
        "cellId": "r2-btl4-35b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2121,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 2
      },
      "startedAt": "2026-10-20T17:41:49.894Z",
      "finishedAt": "2026-10-20T17:47:55.378Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 365484,
        "timing": {
          "startedAt": "2026-10-20T17:41:49.894Z",
          "finishedAt": "2026-10-20T17:47:55.378Z",
          "timeToFirstArtifactMs": 258533,
          "timeToLastArtifactWriteMs": 321653,
          "timeToFirstTokenMs": 252335,
          "firstTurnTtftMs": 8886,
          "timeToFirstToolCallMs": 252821
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 99.82,
        "prefillTokensPerSec": 2428.89,
        "decodeTokens": 7926,
        "peakRssMb": 30544,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T17:57:58.056Z",
        "trialId": "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m",
        "cellId": "r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 7914,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 9,
          "grounding": 3,
          "structure": 9,
          "actionability": 6
        },
        "meanScore": 6.8,
        "justification": "Six consistent APA-style entries retain all seeded figures and useful relevance links. Critical evaluations repeatedly invent unavailable study details: standard-care comparison, consistent site effects, unmeasured confounders, a single-community sample, missing analyses and a privacy checklist. Several author conclusions and causal/significance claims exceed the cards.",
        "artifactSha256": "f19d1ed047a96af9fe103d701378afe1805f792169056d5c91d2ba8b6d67bb40",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm",
      "cell": {
        "id": "r2-btl4-35b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 2
      },
      "startedAt": "2026-10-20T17:48:04.550Z",
      "finishedAt": "2026-10-20T17:54:05.559Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 361009,
        "timing": {
          "startedAt": "2026-10-20T17:48:04.550Z",
          "finishedAt": "2026-10-20T17:54:05.559Z",
          "timeToFirstArtifactMs": 256220,
          "timeToLastArtifactWriteMs": 358535,
          "timeToFirstTokenMs": 248750,
          "firstTurnTtftMs": 8828,
          "timeToFirstToolCallMs": 249402
        },
        "toolCalls": 43,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 106.33,
        "prefillTokensPerSec": 2338.06,
        "decodeTokens": 6152,
        "peakRssMb": 33385,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T17:57:58.057Z",
        "trialId": "craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm",
        "cellId": "r2-btl4-35b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 1459,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 4,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 6.7,
        "justification": "Clear title, summary, metrics table and next actions convert to a retained DOCX. Core numbers are exact, but no confirmation/tracking/next step, status-independent improvements and a must-fix-before-rollout condition are invented. Immediate-photo emails and queue effects add unsupported specifics; coverage is shown as 18 SKUs both before and after without a baseline.",
        "artifactSha256": "7eecf90b53103c73cbc004bc28c6b311a250b136c6d7704e004b10bbdbd3044d",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 2
      },
      "startedAt": "2026-10-20T17:54:21.281Z",
      "finishedAt": "2026-10-20T17:59:18.118Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 296837,
        "timing": {
          "startedAt": "2026-10-20T17:54:21.281Z",
          "finishedAt": "2026-10-20T17:59:18.118Z",
          "timeToFirstArtifactMs": 257067,
          "timeToLastArtifactWriteMs": 291423,
          "timeToFirstTokenMs": 242444,
          "firstTurnTtftMs": 7588,
          "timeToFirstToolCallMs": 243717
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 103.81,
        "prefillTokensPerSec": 2215.51,
        "decodeTokens": 4898,
        "peakRssMb": 27706,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:11:44.355Z",
        "trialId": "fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z",
        "cellId": "r2-btl4-35b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 2
      },
      "startedAt": "2026-10-20T17:59:28.093Z",
      "finishedAt": "2026-10-20T18:03:44.593Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 256500,
        "timing": {
          "startedAt": "2026-10-20T17:59:28.093Z",
          "finishedAt": "2026-10-20T18:03:44.593Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 245208,
          "firstTurnTtftMs": 10563,
          "timeToFirstToolCallMs": 246834
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 109.45,
        "prefillTokensPerSec": 2638.23,
        "decodeTokens": 799,
        "peakRssMb": 27428,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:11:44.356Z",
        "trialId": "dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud",
        "cellId": "r2-btl4-35b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:03:54.147Z",
      "finishedAt": "2026-10-20T18:08:19.751Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265604,
        "timing": {
          "startedAt": "2026-10-20T18:03:54.147Z",
          "finishedAt": "2026-10-20T18:08:19.751Z",
          "timeToFirstArtifactMs": 261201,
          "timeToLastArtifactWriteMs": 261217,
          "timeToFirstTokenMs": 243684,
          "firstTurnTtftMs": 8778,
          "timeToFirstToolCallMs": 248316
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 115.63,
        "prefillTokensPerSec": 2837.07,
        "decodeTokens": 1246,
        "peakRssMb": 27866,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:11:44.356Z",
        "trialId": "interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae",
        "cellId": "r2-btl4-35b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:08:29.491Z",
      "finishedAt": "2026-10-20T18:13:35.798Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 306307,
        "timing": {
          "startedAt": "2026-10-20T18:08:29.491Z",
          "finishedAt": "2026-10-20T18:13:35.798Z",
          "timeToFirstArtifactMs": 279259,
          "timeToLastArtifactWriteMs": 279259,
          "timeToFirstTokenMs": 248388,
          "firstTurnTtftMs": 8632,
          "timeToFirstToolCallMs": 249160
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 105.18,
        "prefillTokensPerSec": 2507.9,
        "decodeTokens": 3046,
        "peakRssMb": 29909,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:28:25.209Z",
        "trialId": "craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb",
        "cellId": "r2-btl4-35b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 2697,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 6
        },
        "meanScore": 6,
        "justification": "Scope/report/verdict sequence completes and catches both SQL injections, debug logging, false-success catch and pagination bug. But required payment/convention reads happen after delivery; an explicitly allowed audit pattern becomes a nit, pre-existing validation is presented as new, and the file-handle leak on export error is missed. Scope also mislabels a commit as unsaved work.",
        "artifactSha256": "67fb58318545b47aadc43f570e8845e9b6d6b1b6e0e2c85a5131f89aaca8f649",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:13:45.578Z",
      "finishedAt": "2026-10-20T18:25:49.091Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 723513,
        "timing": {
          "startedAt": "2026-10-20T18:13:45.578Z",
          "finishedAt": "2026-10-20T18:25:49.091Z",
          "timeToFirstArtifactMs": 310421,
          "timeToLastArtifactWriteMs": 682540,
          "timeToFirstTokenMs": 249113,
          "firstTurnTtftMs": 9093,
          "timeToFirstToolCallMs": 250642
        },
        "toolCalls": 33,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 71.72,
        "prefillTokensPerSec": 1517.97,
        "decodeTokens": 27193,
        "peakRssMb": 32695,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:28:25.209Z",
        "trialId": "craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl",
        "cellId": "r2-btl4-35b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 12901,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 4
        },
        "meanScore": 6,
        "justification": "Completes scope, endpoint review, report and QA/finish, catching retryCount type drift. But it invents a history-to-status rename and implementation 201, misses the undocumented single-widget GET, and flags permitted owner-email omission. Version 1.0 does not establish no existing clients; several breaking design changes are presented as blockers without support.",
        "artifactSha256": "d95730775576721e6a77c71963b2eae7a5abbbb46128fe1a02ed40f17bbdc095",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:26:04.330Z",
      "finishedAt": "2026-10-20T18:31:21.395Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 317065,
        "timing": {
          "startedAt": "2026-10-20T18:26:04.330Z",
          "finishedAt": "2026-10-20T18:31:21.395Z",
          "timeToFirstArtifactMs": 273736,
          "timeToLastArtifactWriteMs": 313020,
          "timeToFirstTokenMs": 244539,
          "firstTurnTtftMs": 9353,
          "timeToFirstToolCallMs": 245782
        },
        "toolCalls": 18,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 103.38,
        "prefillTokensPerSec": 2217.87,
        "decodeTokens": 4308,
        "peakRssMb": 30544,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:42:36.664Z",
        "trialId": "craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55",
        "cellId": "r2-btl4-35b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 5109,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 6,
          "remediation": 8
        },
        "meanScore": 7.3,
        "justification": "Correct source-to-sink paths confirm all three admin defects; safe profile/report patterns are cleared and JSON matches the report. Shell-free execFile, SQL placeholders and requireAdmin are concrete fixes. But guaranteed unauthenticated OS/DB destruction exceeds static evidence, and grouping missing authorization as input validation obscures its distinct root cause.",
        "artifactSha256": "a03111829679a3a42c471469ba723ed28bd05b94ca90cc4f6cc7dd0d33351e31",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:31:33.211Z",
      "finishedAt": "2026-10-20T18:51:50.399Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"large-pr-review:4:targetpr-review.md:frddkaei:rp0:rf0\" stuck for 8m — 32 re-writes (45 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_UNGRADED_AUTHORED_OUTPUT",
            "summary": "An authored output exists without a sniff verdict; quality is capped at the rubric midpoint."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1217188,
        "timing": {
          "startedAt": "2026-10-20T18:31:33.211Z",
          "finishedAt": "2026-10-20T18:51:50.399Z",
          "timeToFirstArtifactMs": 285563,
          "timeToLastArtifactWriteMs": 1157518,
          "timeToFirstTokenMs": 246521,
          "firstTurnTtftMs": 11399,
          "timeToFirstToolCallMs": 247827
        },
        "toolCalls": 71,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 38.72,
        "prefillTokensPerSec": 704.63,
        "decodeTokens": 29402,
        "peakRssMb": 28841,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T18:56:43.441Z",
        "trialId": "large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45",
        "cellId": "r2-btl4-35b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:52:02.200Z",
      "finishedAt": "2026-10-20T18:59:30.819Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 448619,
        "timing": {
          "startedAt": "2026-10-20T18:52:02.200Z",
          "finishedAt": "2026-10-20T18:59:30.819Z",
          "timeToFirstArtifactMs": 274883,
          "timeToLastArtifactWriteMs": 444986,
          "timeToFirstTokenMs": 249228,
          "firstTurnTtftMs": 9086,
          "timeToFirstToolCallMs": 249882
        },
        "toolCalls": 66,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 92.86,
        "prefillTokensPerSec": 2021.71,
        "decodeTokens": 12914,
        "peakRssMb": 33573,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T19:12:42.632Z",
        "trialId": "craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195",
        "cellId": "r2-btl4-35b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 2254,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 7,
          "Invariant discipline": 9,
          "Reconciled verification": 5
        },
        "meanScore": 7,
        "justification": "All seven code sites were listed before edits and changed mechanically; original bodies/tests and changelog survive, with two real 3/3 green runs. But enumeration used file reads rather than real search, all grep attempts failed, and the claimed workspace re-scan is unproven. Counts mix seven occurrences with one excluded file containing two occurrences.",
        "artifactSha256": "0a5b49a4ca07e5bf0b9256369c623412da7ef399636bfd422f928fc3b126351c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 2
      },
      "startedAt": "2026-10-20T18:59:40.381Z",
      "finishedAt": "2026-10-20T19:29:25.335Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 28 while the failing detail kept changing; latest checked file workspace output. Last failure: tests/verify-refactor.mjs did not pass when run with node: exit=1 Command failed: /usr/local/bin/node tests/verify-refactor.mjs node:internal/modules/run_main:107 triggerUncaughtException( ^ AssertionError [ERR_ASSERTION]: formatName(\" ada lovelace \") + actual - expected + 'Ada Lovelace ' - 'Ada Lovelace' ^ at <sandbox>/tests/verify-refactor.mjs:17:10 at ModuleJob.run (node:internal/modules/esm/module_job:439:25) at async node:internal/modules/esm/loader:643:26 at async asyncRunEntryPointWithESMLoader (node:internal/modules/run_main:101:5) { generatedMessage: false, code: 'ERR_ASSERTION', actual: 'Ada Lovelace ', expected: 'Ada Lovelace', operator: 'strictEqual', diff: 'simple' } Node.js v24",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 93.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1784954,
        "timing": {
          "startedAt": "2026-10-20T18:59:40.381Z",
          "finishedAt": "2026-10-20T19:29:25.335Z",
          "timeToFirstArtifactMs": 253570,
          "timeToLastArtifactWriteMs": 1779755,
          "timeToFirstTokenMs": 247896,
          "firstTurnTtftMs": 7906,
          "timeToFirstToolCallMs": 253572
        },
        "toolCalls": 206,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 69.59,
        "prefillTokensPerSec": 1297.07,
        "decodeTokens": 79215,
        "peakRssMb": 40125,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T19:45:35.003Z",
        "trialId": "craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji",
        "cellId": "r2-btl4-35b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 4233,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 1,
          "Reversible-stage execution": 3,
          "Genuine deduplication": 2
        },
        "meanScore": 2,
        "justification": "Source, package and tests were replaced before the first test; no original green baseline exists. Later helper/call-site stages had real green runs, but notes reuse one later receipt and deny deviations. Both sites share a helper, yet signatures, coercion and outputs changed; the final reference check fails on a trailing space.",
        "artifactSha256": "6556efa95dc2d5d0893baf1ac8e03e1f8ff3dba2407c69140e27e15b10419e87",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m",
      "cell": {
        "id": "r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 2
      },
      "startedAt": "2026-10-20T19:29:39.301Z",
      "finishedAt": "2026-10-20T19:41:53.287Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on tasks/1/fix-notes.md; missing signals stayed unchanged: tasks/1/fix-notes.md is missing required content: fix-notes sections in order — nothing in its 1119 bytes matches /^##\\s+Problem[\\s\\S]*^##\\s+Change[\\s\\S]*^##\\s+Files touched[\\s\\S]*^##\\s+Regression coverage[\\s\\S]*^##\\s+Risk[\\s\\S]*^##\\s+How to verify/. Add that content.. Last failure: tasks/1/fix-notes.md is missing required content: fix-notes sections in order — nothing in its 1119 bytes matches /^##\\s+Problem[\\s\\S]*^##\\s+Change[\\s\\S]*^##\\s+Files touched[\\s\\S]*^##\\s+Regression coverage[\\s\\S]*^##\\s+Risk[\\s\\S]*^##\\s+How to verify/. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 96.4% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 733986,
        "timing": {
          "startedAt": "2026-10-20T19:29:39.301Z",
          "finishedAt": "2026-10-20T19:41:53.287Z",
          "timeToFirstArtifactMs": 259989,
          "timeToLastArtifactWriteMs": 732111,
          "timeToFirstTokenMs": 254152,
          "firstTurnTtftMs": 12137,
          "timeToFirstToolCallMs": 255593
        },
        "toolCalls": 76,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 53.65,
        "prefillTokensPerSec": 1162.52,
        "decodeTokens": 19347,
        "peakRssMb": 37005,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T19:46:44.730Z",
        "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m",
        "cellId": "r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1137,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 6,
          "Honest verification": 4
        },
        "meanScore": 6.3,
        "justification": "A real 80-versus-68 red run precedes the one-character threshold fix, and the reproduction plus 9/11 boundaries then pass. However, all three original tests were overwritten, including invalid-quantity coverage. Notes falsely call replacement tests unchanged and omit the test-file edit; review misses this and validation overclaims no regressions.",
        "artifactSha256": "6f66cc4a4d303081cf9baf0c05535e048bf1106828e91bf843d6de7e1d4bb2e9",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 2
      },
      "startedAt": "2026-10-20T19:42:05.979Z",
      "finishedAt": "2026-10-20T19:46:25.126Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 259147,
        "timing": {
          "startedAt": "2026-10-20T19:42:05.979Z",
          "finishedAt": "2026-10-20T19:46:25.126Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 245328,
          "firstTurnTtftMs": 10786,
          "timeToFirstToolCallMs": 247179
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 101.82,
        "prefillTokensPerSec": 2524.21,
        "decodeTokens": 758,
        "peakRssMb": 27441,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T19:47:13.499Z",
        "trialId": "craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 2
      },
      "startedAt": "2026-10-20T19:46:35.714Z",
      "finishedAt": "2026-10-20T20:42:44.501Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 603s — model busy but not delivering (hard digest stuck at 0615ae1600dcd9ee)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3368787,
        "timing": {
          "startedAt": "2026-10-20T19:46:35.714Z",
          "finishedAt": "2026-10-20T20:42:44.501Z",
          "timeToFirstArtifactMs": 3289215,
          "timeToLastArtifactWriteMs": 3289215,
          "timeToFirstTokenMs": 247909,
          "firstTurnTtftMs": 13099,
          "timeToFirstToolCallMs": 248896
        },
        "toolCalls": 36,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 49.34,
        "prefillTokensPerSec": 1090.69,
        "decodeTokens": 147715,
        "peakRssMb": 29182,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T20:57:01.341Z",
        "trialId": "craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 2
      },
      "startedAt": "2026-10-20T20:42:53.951Z",
      "finishedAt": "2026-10-20T20:51:30.168Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7.8,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 516217,
        "timing": {
          "startedAt": "2026-10-20T20:42:53.951Z",
          "finishedAt": "2026-10-20T20:51:30.168Z",
          "timeToFirstArtifactMs": 260420,
          "timeToLastArtifactWriteMs": 510484,
          "timeToFirstTokenMs": 248917,
          "firstTurnTtftMs": 9056,
          "timeToFirstToolCallMs": 249585
        },
        "toolCalls": 73,
        "autoAnswers": 3,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 92.45,
        "prefillTokensPerSec": 1928.02,
        "decodeTokens": 14938,
        "peakRssMb": 36553,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T20:59:38.857Z",
        "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 639,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 9,
          "numbering continuity": 9,
          "professionalism": 2,
          "completeness": 4,
          "print-polish": 6
        },
        "meanScore": 6,
        "justification": "All amounts total 5,415 and numbers 042–044 are continuous; idle client skipped. Books remain at 041/uninvoiced. Two invoices invent November issue dates and give net-30 clients 14-day footer terms; sender/address/contact details are unsupported. All print on one A4 page, but stacked margins waste a large top area.",
        "artifactSha256": "7786c2e88545cf30a2f23c36166a0cdaaed9eb890282dacd556c492d043532a3",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 2
      },
      "startedAt": "2026-10-20T20:51:39.350Z",
      "finishedAt": "2026-10-20T21:17:50.698Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1571348,
        "timing": {
          "startedAt": "2026-10-20T20:51:39.350Z",
          "finishedAt": "2026-10-20T21:17:50.698Z",
          "timeToFirstArtifactMs": 585410,
          "timeToLastArtifactWriteMs": 1566102,
          "timeToFirstTokenMs": 248839,
          "firstTurnTtftMs": 13254,
          "timeToFirstToolCallMs": 249785
        },
        "toolCalls": 62,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 59.23,
        "prefillTokensPerSec": 996.59,
        "decodeTokens": 26628,
        "peakRssMb": 34859,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T21:26:55.283Z",
        "trialId": "craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 2
      },
      "startedAt": "2026-10-20T21:18:04.230Z",
      "finishedAt": "2026-10-20T21:35:12.959Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1028729,
        "timing": {
          "startedAt": "2026-10-20T21:18:04.230Z",
          "finishedAt": "2026-10-20T21:35:12.959Z",
          "timeToFirstArtifactMs": 321405,
          "timeToLastArtifactWriteMs": 733843,
          "timeToFirstTokenMs": 247055,
          "firstTurnTtftMs": 11326,
          "timeToFirstToolCallMs": 248281
        },
        "toolCalls": 61,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 73.55,
        "prefillTokensPerSec": 1190.23,
        "decodeTokens": 12731,
        "peakRssMb": 36911,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T21:42:00.861Z",
        "trialId": "craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 2
      },
      "startedAt": "2026-10-20T21:35:24.529Z",
      "finishedAt": "2026-10-20T22:02:00.044Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-export-generalize:3:targetnone:fr1caszr6:rp0:rf0\" stuck for 8m — 8 re-writes (33 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1595515,
        "timing": {
          "startedAt": "2026-10-20T21:35:24.529Z",
          "finishedAt": "2026-10-20T22:02:00.044Z",
          "timeToFirstArtifactMs": 344516,
          "timeToLastArtifactWriteMs": 1568510,
          "timeToFirstTokenMs": 249014,
          "firstTurnTtftMs": 13353,
          "timeToFirstToolCallMs": 250973
        },
        "toolCalls": 112,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 51.23,
        "prefillTokensPerSec": 815.61,
        "decodeTokens": 50874,
        "peakRssMb": 32343,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T22:11:58.769Z",
        "trialId": "craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 2
      },
      "startedAt": "2026-10-20T22:02:15.867Z",
      "finishedAt": "2026-10-20T22:33:15.705Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1859838,
        "timing": {
          "startedAt": "2026-10-20T22:02:15.867Z",
          "finishedAt": "2026-10-20T22:33:15.705Z",
          "timeToFirstArtifactMs": 1656309,
          "timeToLastArtifactWriteMs": 1854563,
          "timeToFirstTokenMs": 250799,
          "firstTurnTtftMs": 14569,
          "timeToFirstToolCallMs": 253222
        },
        "toolCalls": 179,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 56.42,
        "prefillTokensPerSec": 1058.86,
        "decodeTokens": 45149,
        "peakRssMb": 37638,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T22:42:03.169Z",
        "trialId": "craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 2
      },
      "startedAt": "2026-10-20T22:33:28.656Z",
      "finishedAt": "2026-10-20T22:56:16.831Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a craftbook with a substantive embedded gate script, the report parses, and the task completed",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1368175,
        "timing": {
          "startedAt": "2026-10-20T22:33:28.656Z",
          "finishedAt": "2026-10-20T22:56:16.831Z",
          "timeToFirstArtifactMs": 914797,
          "timeToLastArtifactWriteMs": 1361814,
          "timeToFirstTokenMs": 248373,
          "firstTurnTtftMs": 13125,
          "timeToFirstToolCallMs": 249151
        },
        "toolCalls": 87,
        "autoAnswers": 2,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 48.2,
        "prefillTokensPerSec": 1331.04,
        "decodeTokens": 37242,
        "peakRssMb": 33553,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T22:56:56.598Z",
        "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r",
      "cell": {
        "id": "r2-btl4-35b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 2
      },
      "startedAt": "2026-10-20T22:56:32.309Z",
      "finishedAt": "2026-10-20T23:23:02.125Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:6:targetnone:fr1p9909z:rp0:rf0:m3\" stuck for 8m — 4 re-writes (11 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1589816,
        "timing": {
          "startedAt": "2026-10-20T22:56:32.309Z",
          "finishedAt": "2026-10-20T23:23:02.125Z",
          "timeToFirstArtifactMs": 1097813,
          "timeToLastArtifactWriteMs": 1445889,
          "timeToFirstTokenMs": 273987,
          "firstTurnTtftMs": 13819,
          "timeToFirstToolCallMs": 275689
        },
        "toolCalls": 52,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 52.45,
        "prefillTokensPerSec": 835.99,
        "decodeTokens": 23973,
        "peakRssMb": 32378,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:27:01.115Z",
        "trialId": "craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r",
        "cellId": "r2-btl4-35b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk",
      "cell": {
        "id": "r3-btl4-35b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:23:15.685Z",
      "finishedAt": "2026-10-20T23:28:36.758Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 321073,
        "timing": {
          "startedAt": "2026-10-20T23:23:15.685Z",
          "finishedAt": "2026-10-20T23:28:36.758Z",
          "timeToFirstArtifactMs": 316741,
          "timeToLastArtifactWriteMs": 316750,
          "timeToFirstTokenMs": 286853,
          "firstTurnTtftMs": 9309,
          "timeToFirstToolCallMs": 290461
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 80.71,
        "prefillTokensPerSec": 2739.22,
        "decodeTokens": 1757,
        "peakRssMb": 27783,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:43:50.335Z",
        "trialId": "tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk",
        "cellId": "r3-btl4-35b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 3871,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 8,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "Clean responsive board; all eight win lines, O win, draw, occupied-cell protection, post-win lock and reset work. Winners are highlighted. Empty cells lack accessible names and status has no live region; the visual treatment is a basic dark template with hover effects.",
        "artifactSha256": "dae05a66768ca7c1360be61d206752a4bc84fc3dc35beec57eaf07561be9fa83",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu",
      "cell": {
        "id": "r3-btl4-35b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:29:14.621Z",
      "finishedAt": "2026-10-20T23:35:18.003Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 363382,
        "timing": {
          "startedAt": "2026-10-20T23:29:14.621Z",
          "finishedAt": "2026-10-20T23:35:18.003Z",
          "timeToFirstArtifactMs": 335588,
          "timeToLastArtifactWriteMs": 355769,
          "timeToFirstTokenMs": 290285,
          "firstTurnTtftMs": 11540,
          "timeToFirstToolCallMs": 296659
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 101.92,
        "prefillTokensPerSec": 2391.79,
        "decodeTokens": 4533,
        "peakRssMb": 35274,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:43:50.337Z",
        "trialId": "petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu",
        "cellId": "r3-btl4-35b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/workspace/pet-shop-website/index.html",
        "artifactBytes": 10358,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 5,
          "functionalCompleteness": 7,
          "codeQuality": 6,
          "polish": 4
        },
        "meanScore": 5.5,
        "justification": "HTML precedes the generated local logo; six products, four services and cart totals/reset work. At 375px the fixed cart hides the brand and navigation, with 433px overflow. Product photos depend on unverified remote URLs. No cart removal or live status; checkout is a local demo. Desktop styling is generic.",
        "artifactSha256": "cd1c730ee0a6be5e9a14b82fbdfc9c275a3f71b1b10211c98a66302d5f5c8fa6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb",
      "cell": {
        "id": "r3-btl4-35b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:36:00.616Z",
      "finishedAt": "2026-10-20T23:41:53.926Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 353310,
        "timing": {
          "startedAt": "2026-10-20T23:36:00.616Z",
          "finishedAt": "2026-10-20T23:41:53.926Z",
          "timeToFirstArtifactMs": 333662,
          "timeToLastArtifactWriteMs": 346103,
          "timeToFirstTokenMs": 299889,
          "firstTurnTtftMs": 10650,
          "timeToFirstToolCallMs": 306250
        },
        "toolCalls": 4,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 99.39,
        "prefillTokensPerSec": 2265.78,
        "decodeTokens": 3214,
        "peakRssMb": 28835,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:45:48.521Z",
        "trialId": "tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb",
        "cellId": "r3-btl4-35b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 7098,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 4,
          "functionalCompleteness": 7,
          "codeQuality": 5,
          "polish": 3
        },
        "meanScore": 4.8,
        "justification": "Movement, directional fire, enemy hits, +100 scoring, lives and game-over work locally. Enemies are descending rectangles without return fire. No restart, pause or control instructions; the fixed 800px canvas clips on mobile. Motion/spawn odds are frame-based and gameState refreshes only on keydown. Bare visuals have small particle effects.",
        "artifactSha256": "b2a5be647f7ac1e8b90ea67329728149f391a00f5e78ff99e7d13cac1fd15124",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e",
      "cell": {
        "id": "r3-btl4-35b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:42:32.067Z",
      "finishedAt": "2026-10-20T23:47:49.669Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: riley/63f44c98 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 57.1% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 317602,
        "timing": {
          "startedAt": "2026-10-20T23:42:32.067Z",
          "finishedAt": "2026-10-20T23:47:49.669Z",
          "timeToFirstArtifactMs": 254865,
          "timeToLastArtifactWriteMs": 280124,
          "timeToFirstTokenMs": 253900,
          "firstTurnTtftMs": 7641,
          "timeToFirstToolCallMs": 254867
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 103.61,
        "prefillTokensPerSec": 2133.76,
        "decodeTokens": 2555,
        "peakRssMb": 32233,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:56:59.340Z",
        "trialId": "schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e",
        "cellId": "r3-btl4-35b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j",
      "cell": {
        "id": "r3-btl4-35b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:48:01.372Z",
      "finishedAt": "2026-10-20T23:52:43.709Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 282337,
        "timing": {
          "startedAt": "2026-10-20T23:48:01.372Z",
          "finishedAt": "2026-10-20T23:52:43.709Z",
          "timeToFirstArtifactMs": 261915,
          "timeToLastArtifactWriteMs": 277907,
          "timeToFirstTokenMs": 253748,
          "firstTurnTtftMs": 8132,
          "timeToFirstToolCallMs": 254410
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 112.54,
        "prefillTokensPerSec": 2851.88,
        "decodeTokens": 1355,
        "peakRssMb": 27972,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:56:59.341Z",
        "trialId": "failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j",
        "cellId": "r3-btl4-35b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif",
      "cell": {
        "id": "r3-btl4-35b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:52:55.109Z",
      "finishedAt": "2026-10-20T23:57:12.482Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 257373,
        "timing": {
          "startedAt": "2026-10-20T23:52:55.109Z",
          "finishedAt": "2026-10-20T23:57:12.482Z",
          "timeToFirstArtifactMs": 255702,
          "timeToLastArtifactWriteMs": 255702,
          "timeToFirstTokenMs": 244523,
          "firstTurnTtftMs": 7642,
          "timeToFirstToolCallMs": 245235
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 113.56,
        "prefillTokensPerSec": 2839.82,
        "decodeTokens": 1314,
        "peakRssMb": 27107,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-20T23:57:48.694Z",
        "trialId": "symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif",
        "cellId": "r3-btl4-35b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l",
      "cell": {
        "id": "r3-btl4-35b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 3
      },
      "startedAt": "2026-10-20T23:57:21.432Z",
      "finishedAt": "2026-10-21T00:02:02.611Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 281179,
        "timing": {
          "startedAt": "2026-10-20T23:57:21.432Z",
          "finishedAt": "2026-10-21T00:02:02.611Z",
          "timeToFirstArtifactMs": 273793,
          "timeToLastArtifactWriteMs": 278704,
          "timeToFirstTokenMs": 253165,
          "firstTurnTtftMs": 7729,
          "timeToFirstToolCallMs": 254472
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 107.89,
        "prefillTokensPerSec": 2525.9,
        "decodeTokens": 2515,
        "peakRssMb": 27122,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:12:01.392Z",
        "trialId": "data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l",
        "cellId": "r3-btl4-35b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4",
      "cell": {
        "id": "r3-btl4-35b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:02:14.447Z",
      "finishedAt": "2026-10-21T00:09:32.925Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 438478,
        "timing": {
          "startedAt": "2026-10-21T00:02:14.447Z",
          "finishedAt": "2026-10-21T00:09:32.925Z",
          "timeToFirstArtifactMs": 404835,
          "timeToLastArtifactWriteMs": 435857,
          "timeToFirstTokenMs": 246529,
          "firstTurnTtftMs": 10345,
          "timeToFirstToolCallMs": 249391
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 81.53,
        "prefillTokensPerSec": 1802.54,
        "decodeTokens": 12456,
        "peakRssMb": 30687,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:12:01.393Z",
        "trialId": "incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4",
        "cellId": "r3-btl4-35b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf",
      "cell": {
        "id": "r3-btl4-35b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:09:47.765Z",
      "finishedAt": "2026-10-21T00:14:34.056Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 286291,
        "timing": {
          "startedAt": "2026-10-21T00:09:47.765Z",
          "finishedAt": "2026-10-21T00:14:34.056Z",
          "timeToFirstArtifactMs": 251332,
          "timeToLastArtifactWriteMs": 281407,
          "timeToFirstTokenMs": 248284,
          "firstTurnTtftMs": 12772,
          "timeToFirstToolCallMs": 248793
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.61,
        "prefillTokensPerSec": 2134.62,
        "decodeTokens": 2910,
        "peakRssMb": 27386,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:27:07.592Z",
        "trialId": "ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf",
        "cellId": "r3-btl4-35b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx",
      "cell": {
        "id": "r3-btl4-35b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:14:45.932Z",
      "finishedAt": "2026-10-21T00:19:06.660Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 260728,
        "timing": {
          "startedAt": "2026-10-21T00:14:45.932Z",
          "finishedAt": "2026-10-21T00:19:06.660Z",
          "timeToFirstArtifactMs": 257046,
          "timeToLastArtifactWriteMs": 257046,
          "timeToFirstTokenMs": 242610,
          "firstTurnTtftMs": 7274,
          "timeToFirstToolCallMs": 243633
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 111.15,
        "prefillTokensPerSec": 2913.21,
        "decodeTokens": 1581,
        "peakRssMb": 27068,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:27:07.594Z",
        "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx",
        "cellId": "r3-btl4-35b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7",
      "cell": {
        "id": "r3-btl4-35b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:19:15.681Z",
      "finishedAt": "2026-10-21T00:25:53.682Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/922ebcac exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 84.6% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 398001,
        "timing": {
          "startedAt": "2026-10-21T00:19:15.681Z",
          "finishedAt": "2026-10-21T00:25:53.682Z",
          "timeToFirstArtifactMs": 289591,
          "timeToLastArtifactWriteMs": 377007,
          "timeToFirstTokenMs": 245990,
          "firstTurnTtftMs": 9795,
          "timeToFirstToolCallMs": 247643
        },
        "toolCalls": 35,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 94.22,
        "prefillTokensPerSec": 1972.43,
        "decodeTokens": 8176,
        "peakRssMb": 31513,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:27:07.594Z",
        "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7",
        "cellId": "r3-btl4-35b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:26:05.458Z",
      "finishedAt": "2026-10-21T00:30:32.589Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 267131,
        "timing": {
          "startedAt": "2026-10-21T00:26:05.458Z",
          "finishedAt": "2026-10-21T00:30:32.589Z",
          "timeToFirstArtifactMs": 265348,
          "timeToLastArtifactWriteMs": 265348,
          "timeToFirstTokenMs": 253240,
          "firstTurnTtftMs": 6539,
          "timeToFirstToolCallMs": 255827
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 111.99,
        "prefillTokensPerSec": 2761.88,
        "decodeTokens": 985,
        "peakRssMb": 28243,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:41:59.880Z",
        "trialId": "constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10",
        "cellId": "r3-btl4-35b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:30:43.307Z",
      "finishedAt": "2026-10-21T00:37:58.284Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 434977,
        "timing": {
          "startedAt": "2026-10-21T00:30:43.307Z",
          "finishedAt": "2026-10-21T00:37:58.284Z",
          "timeToFirstArtifactMs": 290283,
          "timeToLastArtifactWriteMs": 433697,
          "timeToFirstTokenMs": 265200,
          "firstTurnTtftMs": 6775,
          "timeToFirstToolCallMs": 266380
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 95.8,
        "prefillTokensPerSec": 2306.15,
        "decodeTokens": 15035,
        "peakRssMb": 28522,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:41:59.881Z",
        "trialId": "records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55",
        "cellId": "r3-btl4-35b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:38:10.674Z",
      "finishedAt": "2026-10-21T00:42:35.571Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 264897,
        "timing": {
          "startedAt": "2026-10-21T00:38:10.674Z",
          "finishedAt": "2026-10-21T00:42:35.571Z",
          "timeToFirstArtifactMs": 262088,
          "timeToLastArtifactWriteMs": 263857,
          "timeToFirstTokenMs": 252660,
          "firstTurnTtftMs": 8278,
          "timeToFirstToolCallMs": 253490
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 113.08,
        "prefillTokensPerSec": 2734.94,
        "decodeTokens": 1192,
        "peakRssMb": 26519,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:57:04.370Z",
        "trialId": "plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da",
        "cellId": "r3-btl4-35b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:42:47.711Z",
      "finishedAt": "2026-10-21T00:49:09.996Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 382285,
        "timing": {
          "startedAt": "2026-10-21T00:42:47.711Z",
          "finishedAt": "2026-10-21T00:49:09.996Z",
          "timeToFirstArtifactMs": 251762,
          "timeToLastArtifactWriteMs": 377742,
          "timeToFirstTokenMs": 244147,
          "firstTurnTtftMs": 8334,
          "timeToFirstToolCallMs": 245165
        },
        "toolCalls": 15,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 113.57,
        "prefillTokensPerSec": 2606.27,
        "decodeTokens": 2672,
        "peakRssMb": 29719,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T00:57:28.629Z",
        "trialId": "meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw",
        "cellId": "r3-btl4-35b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1482,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 9,
          "grounding": 9,
          "actionability": 9,
          "brevity": 9
        },
        "meanScore": 9,
        "justification": "Clearly separates the Sep 14 EU-only overnight-batch decisions, four owned actions, open annual-plan/rollback questions and legal risk. Dates, A2→A1 dependency and all four CSV rows match the transcript/roster. The Aug 30 item correctly schedules the review rather than assigning its meeting date; stale proposals stay rejected. Compact and executable.",
        "artifactSha256": "048704e19149a25757856d1d98acb1ffe019edbcd8cab01f147986ff2a7dfe01",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 3
      },
      "startedAt": "2026-10-21T00:49:21.710Z",
      "finishedAt": "2026-10-21T00:57:02.932Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 461222,
        "timing": {
          "startedAt": "2026-10-21T00:49:21.710Z",
          "finishedAt": "2026-10-21T00:57:02.932Z",
          "timeToFirstArtifactMs": 272745,
          "timeToLastArtifactWriteMs": 387060,
          "timeToFirstTokenMs": 249836,
          "firstTurnTtftMs": 8349,
          "timeToFirstToolCallMs": 250815
        },
        "toolCalls": 12,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 89.1,
        "prefillTokensPerSec": 2202.99,
        "decodeTokens": 15568,
        "peakRssMb": 29684,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T01:01:46.172Z",
        "trialId": "craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89",
        "cellId": "r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 2828,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 4
        },
        "meanScore": 5.5,
        "justification": "Scope, build and evaluate ran in order with source reads and a correct negative-net hand-check. Review nevertheless approves 10-month break-even while losing $2,135 every month. Conservative 9.4% is not 60% of the observed improvement; three displayed drivers are unused, and the scope contradicts its own payback formula. Claimed changed-input verification has no runtime receipt.",
        "artifactSha256": "d2b3c62e9cb27c25063cb96a87b5f0be59222151d3954826cd802529e20f1f43",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 3
      },
      "startedAt": "2026-10-21T01:01:20.536Z",
      "finishedAt": "2026-10-21T01:07:06.493Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on synthesis.md; the same scenario failure stayed unchanged. Last failure: - conflict (launch budget): a superseded value appears outside the Conflicts section; re-read memo-product.md and finance.csv, then keep the historical value only in its reconciliation entry",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 345957,
        "timing": {
          "startedAt": "2026-10-21T01:01:20.536Z",
          "finishedAt": "2026-10-21T01:07:06.493Z",
          "timeToFirstArtifactMs": 271032,
          "timeToLastArtifactWriteMs": 343770,
          "timeToFirstTokenMs": 243957,
          "firstTurnTtftMs": 9233,
          "timeToFirstToolCallMs": 245468
        },
        "toolCalls": 42,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 102.57,
        "prefillTokensPerSec": 2197.6,
        "decodeTokens": 7094,
        "peakRssMb": 33308,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T01:11:52.766Z",
        "trialId": "conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1",
        "cellId": "r3-btl4-35b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 3
      },
      "startedAt": "2026-10-21T01:07:18.395Z",
      "finishedAt": "2026-10-21T01:11:38.977Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 260582,
        "timing": {
          "startedAt": "2026-10-21T01:07:18.395Z",
          "finishedAt": "2026-10-21T01:11:38.977Z",
          "timeToFirstArtifactMs": 258028,
          "timeToLastArtifactWriteMs": 258028,
          "timeToFirstTokenMs": 242449,
          "firstTurnTtftMs": 7366,
          "timeToFirstToolCallMs": 242511
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 111.28,
        "prefillTokensPerSec": 2351.51,
        "decodeTokens": 1468,
        "peakRssMb": 30366,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T01:12:29.593Z",
        "trialId": "docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon",
        "cellId": "r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1039,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 8,
          "candor": 9,
          "usefulness": 8
        },
        "meanScore": 8.3,
        "justification": "All frozen theme values, page counts and the unresolved Callout Emphasis match the tool results; applied style names and preserved content are also tool-supported. The extra claim that content slides collapsed is unverified. The conflict is stated plainly and the Body/accent-box replacement is actionable, although the prediction that even a matched style will look like a leftover is speculative.",
        "artifactSha256": "4ff6297fcc0d9ac9a4b906e1aaac89f9aa242f2d581cb38a9abf54f97b93dfb1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 3
      },
      "startedAt": "2026-10-21T01:11:47.853Z",
      "finishedAt": "2026-10-21T01:18:39.615Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 411762,
        "timing": {
          "startedAt": "2026-10-21T01:11:47.853Z",
          "finishedAt": "2026-10-21T01:18:39.615Z",
          "timeToFirstArtifactMs": 269203,
          "timeToLastArtifactWriteMs": 405830,
          "timeToFirstTokenMs": 248764,
          "firstTurnTtftMs": 8963,
          "timeToFirstToolCallMs": 249208
        },
        "toolCalls": 35,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 97.59,
        "prefillTokensPerSec": 2253.6,
        "decodeTokens": 11885,
        "peakRssMb": 33111,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T01:27:56.430Z",
        "trialId": "craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939",
        "cellId": "r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1303,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 7,
          "convertibility": 9
        },
        "meanScore": 8.5,
        "justification": "Exact source was read and cited before outlining; all seven slide titles match once, in order. Metrics and named actions are correct, but the before-scaling prerequisite and promised handling-error/leak-prevention benefits exceed the brief. The source ledger mislabels internal search as a zero-result web search. Concise H1 slide blocks and real conversion/save/copy support convertibility.",
        "artifactSha256": "09e678589d2358817464e57ba3138153608ebebe3905139509bb9817de4c3781",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 3
      },
      "startedAt": "2026-10-21T01:18:54.868Z",
      "finishedAt": "2026-10-21T01:31:23.605Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 748737,
        "timing": {
          "startedAt": "2026-10-21T01:18:54.868Z",
          "finishedAt": "2026-10-21T01:31:23.605Z",
          "timeToFirstArtifactMs": 362282,
          "timeToLastArtifactWriteMs": 743455,
          "timeToFirstTokenMs": 241287,
          "firstTurnTtftMs": 6282,
          "timeToFirstToolCallMs": 248731
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 88.99,
        "prefillTokensPerSec": 1921.69,
        "decodeTokens": 19324,
        "peakRssMb": 30710,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T01:42:32.440Z",
        "trialId": "wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt",
        "cellId": "r3-btl4-35b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5607,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 3,
          "citationQuality": 6,
          "calibration": 4
        },
        "meanScore": 4.3,
        "justification": "Core chronology and the disputed label are accurate and cited, but the brief mainly repeats source summaries. It adds unsupported symbolic-computing foresight, manufacturing/funding causes, earlier Babbage methods and terminology history. Never completed in their lifetimes becomes never built and an unjustified cannot-verify claim. Adjacent source IDs do not support these added assertions.",
        "artifactSha256": "55d80187ed03c47c53e279aa928be07ec735a89911e7102c53863328c01806af",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 3
      },
      "startedAt": "2026-10-21T01:31:39.106Z",
      "finishedAt": "2026-10-21T03:05:37.557Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "hit hard ceiling (2700000ms + 45m progress extensions) — last hard progress 5s before the end; the deliverable never closed",
        "failureMode": "timeout",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 1.2,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 1.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 16.7% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 5638451,
        "timing": {
          "startedAt": "2026-10-21T01:31:39.106Z",
          "finishedAt": "2026-10-21T03:05:37.557Z",
          "timeToFirstArtifactMs": 704612,
          "timeToLastArtifactWriteMs": 5637317,
          "timeToFirstTokenMs": 250460,
          "firstTurnTtftMs": 10414,
          "timeToFirstToolCallMs": 252645
        },
        "toolCalls": 545,
        "autoAnswers": 11,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 72.07,
        "prefillTokensPerSec": 1571.53,
        "decodeTokens": 292460,
        "peakRssMb": 37467,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:14:36.287Z",
        "trialId": "craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4",
        "cellId": "r3-btl4-35b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 3190,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 1,
          "prioritization": 3,
          "conflict-handling": 1,
          "realism": 2
        },
        "meanScore": 1.8,
        "justification": "The requested-path report invents meetings and times, omitting Northwind, budget/Priya and All-hands. Five focus blocks largely fit real gaps and follow-up dates are usable, but priorities depend on fictional events and FU-11 timing contradicts itself. Duplicate migration syncs replace the real budget/Priya conflict; prep and load claims are unreliable.",
        "artifactSha256": "0798c5efdbeaaea79b3117cdba0c7240885ded1047e85f5596f8bc4f9eb6756e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:05:56.471Z",
      "finishedAt": "2026-10-21T03:16:24.094Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 627623,
        "timing": {
          "startedAt": "2026-10-21T03:05:56.471Z",
          "finishedAt": "2026-10-21T03:16:24.094Z",
          "timeToFirstArtifactMs": 276506,
          "timeToLastArtifactWriteMs": 622096,
          "timeToFirstTokenMs": 250031,
          "firstTurnTtftMs": 10127,
          "timeToFirstToolCallMs": 250799
        },
        "toolCalls": 59,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 79.87,
        "prefillTokensPerSec": 1936.54,
        "decodeTokens": 14456,
        "peakRssMb": 33655,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:26:59.470Z",
        "trialId": "craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3",
        "cellId": "r3-btl4-35b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 2432,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:16:33.474Z",
      "finishedAt": "2026-10-21T03:26:30.867Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 597393,
        "timing": {
          "startedAt": "2026-10-21T03:16:33.474Z",
          "finishedAt": "2026-10-21T03:26:30.867Z",
          "timeToFirstArtifactMs": 259123,
          "timeToLastArtifactWriteMs": 539878,
          "timeToFirstTokenMs": 252286,
          "firstTurnTtftMs": 8379,
          "timeToFirstToolCallMs": 252771
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 80.16,
        "prefillTokensPerSec": 1910.39,
        "decodeTokens": 22980,
        "peakRssMb": 32634,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:27:49.314Z",
        "trialId": "craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l",
        "cellId": "r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 8132,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 9,
          "grounding": 3,
          "structure": 9,
          "actionability": 6
        },
        "meanScore": 6.8,
        "justification": "All six sources, dates and quantitative results match the cards; APA-style entries and the three annotations are uniform. Yet it invents peer review, comparison baselines, cross-sectional design, study conclusions, remote-coaching evidence, model validation gaps and privacy-regulation history. Relevance suggestions are useful but generic and sometimes rest on those unsupported claims.",
        "artifactSha256": "68f1a276b41cae545219373a4d9f43693e74ad9db6762f9cc5f8813c317656ef",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho",
      "cell": {
        "id": "r3-btl4-35b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:26:40.474Z",
      "finishedAt": "2026-10-21T03:34:12.520Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on workspace output; missing signals stayed unchanged: tasks/eval/report.docx is not a real ZIP container (it is text starting \"<!DOCTYPE html> <html> <head> <meta char\"). Produce it through the DocBlocks conversion route rather than writing text to the binary path., mock docblocks: MCP tool \"save_artifact\" was never called; writing about it does not count — \"save_artifact\" is on your tool roster, call it like any other tool and use its response in the deliverable. Last failure: tasks/eval/report.docx is not a real ZIP container (it is text starting \"<!DOCTYPE html> <html> <head> <meta char\"). Produce it through the DocBlocks conversion route rather than writing text to the binary path.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.2,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 452046,
        "timing": {
          "startedAt": "2026-10-21T03:26:40.474Z",
          "finishedAt": "2026-10-21T03:34:12.520Z",
          "timeToFirstArtifactMs": 257134,
          "timeToLastArtifactWriteMs": 447613,
          "timeToFirstTokenMs": 248520,
          "firstTurnTtftMs": 8299,
          "timeToFirstToolCallMs": 249161
        },
        "toolCalls": 62,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 99.27,
        "prefillTokensPerSec": 2248.46,
        "decodeTokens": 12822,
        "peakRssMb": 37288,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:43:51.774Z",
        "trialId": "craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho",
        "cellId": "r3-btl4-35b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 2743,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 3,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 6.3,
        "justification": "Numbers and action names match, but the self-service workflow, routing, tracking, causes, owners and scale-readiness claims are invented. Clear title, summary, metrics/action tables and references make the Markdown structurally convertible. The produced .docx is HTML and save_artifact was omitted; the fixed trial correctly records failure.",
        "artifactSha256": "41b562590228f0a737d9df6e1655fb1a9fc1a9cc4ec9bccf4276273de5c79f1c",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:34:27.945Z",
      "finishedAt": "2026-10-21T03:38:55.302Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 267357,
        "timing": {
          "startedAt": "2026-10-21T03:34:27.945Z",
          "finishedAt": "2026-10-21T03:38:55.302Z",
          "timeToFirstArtifactMs": 260902,
          "timeToLastArtifactWriteMs": 260935,
          "timeToFirstTokenMs": 241689,
          "firstTurnTtftMs": 6557,
          "timeToFirstToolCallMs": 242965
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 109.09,
        "prefillTokensPerSec": 2686.31,
        "decodeTokens": 2186,
        "peakRssMb": 27779,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:41:54.960Z",
        "trialId": "fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5",
        "cellId": "r3-btl4-35b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:39:04.534Z",
      "finishedAt": "2026-10-21T03:43:19.994Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 255460,
        "timing": {
          "startedAt": "2026-10-21T03:39:04.534Z",
          "finishedAt": "2026-10-21T03:43:19.994Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 245024,
          "firstTurnTtftMs": 10070,
          "timeToFirstToolCallMs": 246958
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 109.64,
        "prefillTokensPerSec": 2720.37,
        "decodeTokens": 778,
        "peakRssMb": 27367,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:44:29.969Z",
        "trialId": "dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4",
        "cellId": "r3-btl4-35b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:43:28.810Z",
      "finishedAt": "2026-10-21T03:47:54.149Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 5 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265339,
        "timing": {
          "startedAt": "2026-10-21T03:43:28.810Z",
          "finishedAt": "2026-10-21T03:47:54.149Z",
          "timeToFirstArtifactMs": 254375,
          "timeToLastArtifactWriteMs": 260606,
          "timeToFirstTokenMs": 242586,
          "firstTurnTtftMs": 7755,
          "timeToFirstToolCallMs": 247713
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 111.84,
        "prefillTokensPerSec": 2760.8,
        "decodeTokens": 1766,
        "peakRssMb": 28093,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:56:55.968Z",
        "trialId": "interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr",
        "cellId": "r3-btl4-35b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:48:03.012Z",
      "finishedAt": "2026-10-21T03:53:09.350Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 306338,
        "timing": {
          "startedAt": "2026-10-21T03:48:03.012Z",
          "finishedAt": "2026-10-21T03:53:09.350Z",
          "timeToFirstArtifactMs": 282577,
          "timeToLastArtifactWriteMs": 282577,
          "timeToFirstTokenMs": 247366,
          "firstTurnTtftMs": 7602,
          "timeToFirstToolCallMs": 248508
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 105.43,
        "prefillTokensPerSec": 2454.51,
        "decodeTokens": 3626,
        "peakRssMb": 29015,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T03:57:55.718Z",
        "trialId": "craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk",
        "cellId": "r3-btl4-35b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 4259,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 6,
          "usefulness": 6
        },
        "meanScore": 6,
        "justification": "Uses snapshot scope, a cited severity table, artifact reread and verdict, but reads all mandated workspace inputs only after submitting the report. Correctly catches both SQL sinks, debug logging, settings and pagination. Export advice misses guaranteed cleanup, while audit, null-input and error-logging warnings lack contract support; some payment issues predate the diff.",
        "artifactSha256": "700e6469bc13b47106d4d4f763edfbcc0aa9a074d6e326ec2d9c3dce8fb1a3d2",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 3
      },
      "startedAt": "2026-10-21T03:53:18.973Z",
      "finishedAt": "2026-10-21T04:03:22.045Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 603072,
        "timing": {
          "startedAt": "2026-10-21T03:53:18.973Z",
          "finishedAt": "2026-10-21T04:03:22.045Z",
          "timeToFirstArtifactMs": 267633,
          "timeToLastArtifactWriteMs": 543675,
          "timeToFirstTokenMs": 247788,
          "firstTurnTtftMs": 8048,
          "timeToFirstToolCallMs": 248543
        },
        "toolCalls": 35,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 79.38,
        "prefillTokensPerSec": 1875.92,
        "decodeTokens": 23449,
        "peakRssMb": 31953,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T04:13:08.346Z",
        "trialId": "craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl",
        "cellId": "r3-btl4-35b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 14418,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 4
        },
        "meanScore": 5.5,
        "justification": "Scope, endpoint review and gated report were produced, but contract conventions were read only after reporting. Useful retryCount and design fixes coexist with a false optional-nullable finding, missed unimplemented history and undocumented widget GET, and an invented no-prior-version assumption. The summary counts six failed criteria while its own table marks seven.",
        "artifactSha256": "15a34d032a326bfeccfe2856aa219a08068678d7d517ed13ffefbaef526c43ef",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 3
      },
      "startedAt": "2026-10-21T04:03:32.361Z",
      "finishedAt": "2026-10-21T04:08:39.320Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 306959,
        "timing": {
          "startedAt": "2026-10-21T04:03:32.361Z",
          "finishedAt": "2026-10-21T04:08:39.320Z",
          "timeToFirstArtifactMs": 267687,
          "timeToLastArtifactWriteMs": 303824,
          "timeToFirstTokenMs": 243630,
          "firstTurnTtftMs": 8328,
          "timeToFirstToolCallMs": 244875
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.44,
        "prefillTokensPerSec": 2390.24,
        "decodeTokens": 4687,
        "peakRssMb": 28478,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T04:13:08.348Z",
        "trialId": "craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev",
        "cellId": "r3-btl4-35b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 4642,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 6,
          "remediation": 8
        },
        "meanScore": 7.3,
        "justification": "Correctly traces both admin request inputs to shell/SQL sinks and confirms the missing admin guard; all deliberate safe patterns remain unflagged. Report and JSON give concrete matching fixes. The blast-radius claim of internet-wide unauthenticated access and whole-database modification exceeds the supplied routing/deployment/privilege evidence, despite later caveats.",
        "artifactSha256": "4afa2b2a4dd6989c7f70e1b2cb4907d1ca6539e4e260388823b286a9ef19a772",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 3
      },
      "startedAt": "2026-10-21T04:08:48.628Z",
      "finishedAt": "2026-10-21T04:49:16.138Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"large-pr-review:5:targetpr-review.md:fr1i0sit8:rp0:rf0\" stuck for 15m — 3 re-writes (8 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_UNGRADED_AUTHORED_OUTPUT",
            "summary": "An authored output exists without a sniff verdict; quality is capped at the rubric midpoint."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2427510,
        "timing": {
          "startedAt": "2026-10-21T04:08:48.628Z",
          "finishedAt": "2026-10-21T04:49:16.138Z",
          "timeToFirstArtifactMs": 580070,
          "timeToLastArtifactWriteMs": 2419494,
          "timeToFirstTokenMs": 245772,
          "firstTurnTtftMs": 10360,
          "timeToFirstToolCallMs": 247160
        },
        "toolCalls": 62,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 35.95,
        "prefillTokensPerSec": 647.12,
        "decodeTokens": 69257,
        "peakRssMb": 28910,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T04:57:11.960Z",
        "trialId": "large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg",
        "cellId": "r3-btl4-35b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 3
      },
      "startedAt": "2026-10-21T04:49:25.600Z",
      "finishedAt": "2026-10-21T05:11:26.808Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1321208,
        "timing": {
          "startedAt": "2026-10-21T04:49:25.600Z",
          "finishedAt": "2026-10-21T05:11:26.808Z",
          "timeToFirstArtifactMs": 272638,
          "timeToLastArtifactWriteMs": 1318604,
          "timeToFirstTokenMs": 251948,
          "firstTurnTtftMs": 8027,
          "timeToFirstToolCallMs": 252629
        },
        "toolCalls": 88,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 72.55,
        "prefillTokensPerSec": 1735.46,
        "decodeTokens": 72080,
        "peakRssMb": 36098,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T05:13:35.597Z",
        "trialId": "craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m",
        "cellId": "r3-btl4-35b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 2419,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 8,
          "Invariant discipline": 9,
          "Reconciled verification": 4
        },
        "meanScore": 7,
        "justification": "A real pre-edit grep found all 7 code sites plus 4 excluded mentions. Code/test edits are exactly the rename and tests genuinely pass. Verification claims a repository re-scan, but only artifact greps and manual reads are captured afterward. Its 8=7+1 ledger mixes sites and files, omits changelog occurrences, and falsely says loadUser is absent. The quoted green-suite output is genuine.",
        "artifactSha256": "d3537b560d3633acad62047d5b59345f62ad51363f5317756892e0883e8df82a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 3
      },
      "startedAt": "2026-10-21T05:11:38.324Z",
      "finishedAt": "2026-10-21T05:29:56.187Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1097863,
        "timing": {
          "startedAt": "2026-10-21T05:11:38.324Z",
          "finishedAt": "2026-10-21T05:29:56.187Z",
          "timeToFirstArtifactMs": 249086,
          "timeToLastArtifactWriteMs": 1092482,
          "timeToFirstTokenMs": 247237,
          "firstTurnTtftMs": 7405,
          "timeToFirstToolCallMs": 249088
        },
        "toolCalls": 139,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 77.3,
        "prefillTokensPerSec": 1620.01,
        "decodeTokens": 52732,
        "peakRssMb": 36410,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T05:43:58.089Z",
        "trialId": "craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267",
        "cellId": "r3-btl4-35b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 1827,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 6,
          "Reversible-stage execution": 8,
          "Genuine deduplication": 9
        },
        "meanScore": 7.7,
        "justification": "Original 2 tests and 9 added cases were green before either caller changed; all three real stages stayed 11/11 green and final logic matches the originals. But a speculative shared module preceded inputs/baseline and is omitted from deviations; baseline miscounts added tests, and claimed post-edit grep/wc runs lack capture evidence.",
        "artifactSha256": "3c05de528f3f744be21ace8710765b16df06ff0a586ce6d9a6c0f144124a7330",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps",
      "cell": {
        "id": "r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 3
      },
      "startedAt": "2026-10-21T05:30:10.596Z",
      "finishedAt": "2026-10-21T05:37:09.111Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 418515,
        "timing": {
          "startedAt": "2026-10-21T05:30:10.596Z",
          "finishedAt": "2026-10-21T05:37:09.111Z",
          "timeToFirstArtifactMs": 252561,
          "timeToLastArtifactWriteMs": 414506,
          "timeToFirstTokenMs": 249221,
          "firstTurnTtftMs": 8865,
          "timeToFirstToolCallMs": 250436
        },
        "toolCalls": 57,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 101.01,
        "prefillTokensPerSec": 2294.65,
        "decodeTokens": 11439,
        "peakRssMb": 34365,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T05:43:58.090Z",
        "trialId": "craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps",
        "cellId": "r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 1663,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 9,
          "Honest verification": 8
        },
        "meanScore": 8.7,
        "justification": "A real 80 != 68 red run preceded the one-character >= fix; the original three tests were restored before that run and survived in the final 4/4 green suite. Validation quotes the actual run. Fix notes omit the test-file edit and retain an outdated unverified-suite caveat; review accepts these without reconciling the complete change history.",
        "artifactSha256": "c2c1e638f4446e06081d8c5ccc05f4388934e1f2ba80bb23e097d383c19ecf96",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 3
      },
      "startedAt": "2026-10-21T05:37:22.721Z",
      "finishedAt": "2026-10-21T05:41:36.249Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 253528,
        "timing": {
          "startedAt": "2026-10-21T05:37:22.721Z",
          "finishedAt": "2026-10-21T05:41:36.249Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 244668,
          "firstTurnTtftMs": 9588,
          "timeToFirstToolCallMs": 246309
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 109.37,
        "prefillTokensPerSec": 2744.83,
        "decodeTokens": 417,
        "peakRssMb": 27308,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T05:42:01.627Z",
        "trialId": "craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 3
      },
      "startedAt": "2026-10-21T05:41:45.310Z",
      "finishedAt": "2026-10-21T06:02:31.353Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on task quarterly-brief/2 craftbook; missing signals stayed unchanged: task quarterly-brief/2 still carries the original failing step prompt verbatim — read the task's craftbook, find what the final gate demands, and rewrite that step's prompt to instruct it, task quarterly-brief/2's craftbook still has 2 steps — add a verification step at the end of the recipe, task quarterly-brief/2 has status \"active\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.. Last failure: task quarterly-brief/2 still carries the original failing step prompt verbatim — read the task's craftbook, find what the final gate demands, and rewrite that step's prompt to instruct it",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 1.1,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 2.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 25% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1246043,
        "timing": {
          "startedAt": "2026-10-21T05:41:45.310Z",
          "finishedAt": "2026-10-21T06:02:31.353Z",
          "timeToFirstArtifactMs": 836498,
          "timeToLastArtifactWriteMs": 1213256,
          "timeToFirstTokenMs": 247213,
          "firstTurnTtftMs": 12022,
          "timeToFirstToolCallMs": 247681
        },
        "toolCalls": 84,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 70.03,
        "prefillTokensPerSec": 1524.39,
        "decodeTokens": 62947,
        "peakRssMb": 36375,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T06:11:56.919Z",
        "trialId": "craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 3
      },
      "startedAt": "2026-10-21T06:02:42.288Z",
      "finishedAt": "2026-10-21T06:12:09.084Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 566796,
        "timing": {
          "startedAt": "2026-10-21T06:02:42.288Z",
          "finishedAt": "2026-10-21T06:12:09.084Z",
          "timeToFirstArtifactMs": 260482,
          "timeToLastArtifactWriteMs": 556868,
          "timeToFirstTokenMs": 248707,
          "firstTurnTtftMs": 8558,
          "timeToFirstToolCallMs": 249372
        },
        "toolCalls": 63,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 84.35,
        "prefillTokensPerSec": 2129.13,
        "decodeTokens": 20113,
        "peakRssMb": 34883,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T06:29:17.042Z",
        "trialId": "craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 575,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 9,
          "numbering continuity": 9,
          "professionalism": 2,
          "completeness": 4,
          "print-polish": 3
        },
        "meanScore": 5.4,
        "justification": "Ledger amounts reconcile to 5,415 and 042-044 are contiguous; idle client skipped. Books still say 041/uninvoiced. Report claims sent without evidence; 042 lacks a visible number/date, 043/044 violate terms and invent contacts. Print review shows 2/2/1 pages: excessive gaps split payment details and totals.",
        "artifactSha256": "9ca49e913df1a1a24e6a5949b0176cd04bb6e99853f156ab6f114fa0066337d5",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 3
      },
      "startedAt": "2026-10-21T06:12:20.357Z",
      "finishedAt": "2026-10-21T06:55:58.330Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2617973,
        "timing": {
          "startedAt": "2026-10-21T06:12:20.357Z",
          "finishedAt": "2026-10-21T06:55:58.330Z",
          "timeToFirstArtifactMs": 430687,
          "timeToLastArtifactWriteMs": 2612554,
          "timeToFirstTokenMs": 247645,
          "firstTurnTtftMs": 12155,
          "timeToFirstToolCallMs": 248665
        },
        "toolCalls": 113,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 72.39,
        "prefillTokensPerSec": 1270.54,
        "decodeTokens": 42204,
        "peakRssMb": 36273,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T06:57:20.498Z",
        "trialId": "craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 3
      },
      "startedAt": "2026-10-21T06:56:10.424Z",
      "finishedAt": "2026-10-21T07:03:22.584Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 432160,
        "timing": {
          "startedAt": "2026-10-21T06:56:10.424Z",
          "finishedAt": "2026-10-21T07:03:22.584Z",
          "timeToFirstArtifactMs": 303862,
          "timeToLastArtifactWriteMs": 380517,
          "timeToFirstTokenMs": 245944,
          "firstTurnTtftMs": 10072,
          "timeToFirstToolCallMs": 247961
        },
        "toolCalls": 47,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 100.4,
        "prefillTokensPerSec": 2009.69,
        "decodeTokens": 12650,
        "peakRssMb": 34327,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T07:12:16.560Z",
        "trialId": "craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 3
      },
      "startedAt": "2026-10-21T07:03:38.329Z",
      "finishedAt": "2026-10-21T07:15:53.337Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: esra/cfc78d0a exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `run_installed_script` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "task-paused-unrecovered"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 4.9,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 6.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 66.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 735008,
        "timing": {
          "startedAt": "2026-10-21T07:03:38.329Z",
          "finishedAt": "2026-10-21T07:15:53.337Z",
          "timeToFirstArtifactMs": 296870,
          "timeToLastArtifactWriteMs": 715138,
          "timeToFirstTokenMs": 247283,
          "firstTurnTtftMs": 12198,
          "timeToFirstToolCallMs": 249715
        },
        "toolCalls": 50,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 71.98,
        "prefillTokensPerSec": 1521.16,
        "decodeTokens": 30568,
        "peakRssMb": 31500,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T07:27:15.671Z",
        "trialId": "craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 3
      },
      "startedAt": "2026-10-21T07:16:03.166Z",
      "finishedAt": "2026-10-21T07:46:11.629Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "no real progress for 604s — model busy but not delivering (hard digest stuck at fc1b8c823027b5ce)",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.1,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 5.7,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 57.1% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1808463,
        "timing": {
          "startedAt": "2026-10-21T07:16:03.166Z",
          "finishedAt": "2026-10-21T07:46:11.629Z",
          "timeToFirstArtifactMs": 1628686,
          "timeToLastArtifactWriteMs": 1628686,
          "timeToFirstTokenMs": 248214,
          "firstTurnTtftMs": 12599,
          "timeToFirstToolCallMs": 249026
        },
        "toolCalls": 23,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 63.12,
        "prefillTokensPerSec": 1482.61,
        "decodeTokens": 85960,
        "peakRssMb": 29203,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T07:57:18.443Z",
        "trialId": "craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 3
      },
      "startedAt": "2026-10-21T07:46:21.341Z",
      "finishedAt": "2026-10-21T08:48:43.832Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: elin/4eafd4b1 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `run_installed_script` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.7,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3742491,
        "timing": {
          "startedAt": "2026-10-21T07:46:21.341Z",
          "finishedAt": "2026-10-21T08:48:43.832Z",
          "timeToFirstArtifactMs": 3607539,
          "timeToLastArtifactWriteMs": 3624464,
          "timeToFirstTokenMs": 250841,
          "firstTurnTtftMs": 12085,
          "timeToFirstToolCallMs": 254429
        },
        "toolCalls": 156,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 48.8,
        "prefillTokensPerSec": 896.69,
        "decodeTokens": 144527,
        "peakRssMb": 37271,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T08:57:19.426Z",
        "trialId": "craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh",
      "cell": {
        "id": "r3-btl4-35b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "btl4-35b-q4",
        "targetId": "btl4-35b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 3
      },
      "startedAt": "2026-10-21T08:48:54.736Z",
      "finishedAt": "2026-10-21T09:49:05.601Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-16-m5max-local-seven-completed-arms",
      "outcome": {
        "success": false,
        "reason": "retry loop (fast-path): sniff \"craftbook-author-fanout:7:targetnone:fr1tjwffw:rp0:rf0:m4\" stuck for 17m — 4 re-writes (31 tool calls) without sniff movement. Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.4,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 87.5% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3610865,
        "timing": {
          "startedAt": "2026-10-21T08:48:54.736Z",
          "finishedAt": "2026-10-21T09:49:05.601Z",
          "timeToFirstArtifactMs": 592600,
          "timeToLastArtifactWriteMs": 3604239,
          "timeToFirstTokenMs": 248282,
          "firstTurnTtftMs": 12720,
          "timeToFirstToolCallMs": 250270
        },
        "toolCalls": 126,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "btl4-35b-q4"
        },
        "decodeTokensPerSec": 59.19,
        "prefillTokensPerSec": 1155.98,
        "decodeTokens": 75808,
        "peakRssMb": 34218,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T09:57:35.859Z",
        "trialId": "craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh",
        "cellId": "r3-btl4-35b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "btl4-35b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 1
      },
      "startedAt": "2026-10-21T09:53:34.651Z",
      "finishedAt": "2026-10-21T09:58:06.800Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 272149,
        "timing": {
          "startedAt": "2026-10-21T09:53:34.651Z",
          "finishedAt": "2026-10-21T09:58:06.800Z",
          "timeToFirstArtifactMs": 268306,
          "timeToLastArtifactWriteMs": 268310,
          "timeToFirstTokenMs": 247563,
          "firstTurnTtftMs": 8342,
          "timeToFirstToolCallMs": 250263
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 171.99,
        "prefillTokensPerSec": 2774.2,
        "decodeTokens": 2855,
        "peakRssMb": 29156,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:16:31.792Z",
        "trialId": "tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 6672,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 9,
          "codeQuality": 7,
          "polish": 7
        },
        "meanScore": 7.3,
        "justification": "All eight X lines, O win, draw, occupied-cell protection, terminal locking and reset pass; persistent scores and win pulses add polish. The board fits mobile. Styling is basic; static cell labels omit current marks, the status lacks live announcements, and animation ignores reduced motion.",
        "artifactSha256": "5964c395d3175d8fd0b9e2dc9b550f7c8d270f8fb49b9882a2ba064ff0f76c3b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 1
      },
      "startedAt": "2026-10-21T09:58:42.121Z",
      "finishedAt": "2026-10-21T10:04:40.326Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 358205,
        "timing": {
          "startedAt": "2026-10-21T09:58:42.121Z",
          "finishedAt": "2026-10-21T10:04:40.326Z",
          "timeToFirstArtifactMs": 295347,
          "timeToLastArtifactWriteMs": 353167,
          "timeToFirstTokenMs": 247581,
          "firstTurnTtftMs": 8366,
          "timeToFirstToolCallMs": 268899
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 131.52,
        "prefillTokensPerSec": 2282.21,
        "decodeTokens": 11922,
        "peakRssMb": 35215,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:16:31.792Z",
        "trialId": "petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/workspace/pet-shop-website/index.html",
        "artifactBytes": 20080,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 7,
          "functionalCompleteness": 7,
          "codeQuality": 7,
          "polish": 6
        },
        "meanScore": 6.8,
        "justification": "Readable desktop layout, generated local PNG, labeled demo form and responsive navigation work. Capture shows image generation before HTML, violating the requested order. Mobile hero shrinks to 80px and clips its artwork; menu aria-expanded stays true after closing. Emoji/template content limits polish.",
        "artifactSha256": "95235b149d70756705cdb29695d5f041e86e59854f6ccb5ce8033f047f42da98",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:05:21.340Z",
      "finishedAt": "2026-10-21T10:10:09.266Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 287926,
        "timing": {
          "startedAt": "2026-10-21T10:05:21.340Z",
          "finishedAt": "2026-10-21T10:10:09.266Z",
          "timeToFirstArtifactMs": 282628,
          "timeToLastArtifactWriteMs": 282632,
          "timeToFirstTokenMs": 247821,
          "firstTurnTtftMs": 8463,
          "timeToFirstToolCallMs": 252102
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 155.2,
        "prefillTokensPerSec": 2747.45,
        "decodeTokens": 4965,
        "peakRssMb": 29353,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:16:31.793Z",
        "trialId": "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 11428,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 2,
          "codeQuality": 3,
          "polish": 4
        },
        "meanScore": 3.8,
        "justification": "Recognizable tanks, HUD, particles and restart are present, but player.isPlayer is never set: shots count as hostile, hit the player and cannot damage enemies. Unmodified real-time replay loses all three lives within 1.1s of firing, at score zero. HUD stays at one life on game over; fixed canvas clips on mobile.",
        "artifactSha256": "3203b976bee753b259bf86a903ff23ebd29463b4a7bb409963c36d81d8a872ea",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:10:45.269Z",
      "finishedAt": "2026-10-21T10:16:11.331Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 326062,
        "timing": {
          "startedAt": "2026-10-21T10:10:45.269Z",
          "finishedAt": "2026-10-21T10:16:11.331Z",
          "timeToFirstArtifactMs": 256588,
          "timeToLastArtifactWriteMs": 320890,
          "timeToFirstTokenMs": 245648,
          "firstTurnTtftMs": 6619,
          "timeToFirstToolCallMs": 256590
        },
        "toolCalls": 17,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 162.28,
        "prefillTokensPerSec": 2307.91,
        "decodeTokens": 4273,
        "peakRssMb": 33314,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:17:21.408Z",
        "trialId": "schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:16:21.249Z",
      "finishedAt": "2026-10-21T10:21:38.699Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 317450,
        "timing": {
          "startedAt": "2026-10-21T10:16:21.249Z",
          "finishedAt": "2026-10-21T10:21:38.699Z",
          "timeToFirstArtifactMs": 251687,
          "timeToLastArtifactWriteMs": 314314,
          "timeToFirstTokenMs": 246266,
          "firstTurnTtftMs": 7157,
          "timeToFirstToolCallMs": 246965
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 169.63,
        "prefillTokensPerSec": 2144.47,
        "decodeTokens": 2051,
        "peakRssMb": 35946,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:27:19.420Z",
        "trialId": "failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:21:48.204Z",
      "finishedAt": "2026-10-21T10:26:27.808Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 279604,
        "timing": {
          "startedAt": "2026-10-21T10:21:48.204Z",
          "finishedAt": "2026-10-21T10:26:27.808Z",
          "timeToFirstArtifactMs": 276497,
          "timeToLastArtifactWriteMs": 276497,
          "timeToFirstTokenMs": 246311,
          "firstTurnTtftMs": 7282,
          "timeToFirstToolCallMs": 246850
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 134.77,
        "prefillTokensPerSec": 2696.81,
        "decodeTokens": 3987,
        "peakRssMb": 28354,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:27:19.421Z",
        "trialId": "symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:26:37.717Z",
      "finishedAt": "2026-10-21T10:31:31.765Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 294048,
        "timing": {
          "startedAt": "2026-10-21T10:26:37.717Z",
          "finishedAt": "2026-10-21T10:31:31.765Z",
          "timeToFirstArtifactMs": 283527,
          "timeToLastArtifactWriteMs": 285744,
          "timeToFirstTokenMs": 245737,
          "firstTurnTtftMs": 7335,
          "timeToFirstToolCallMs": 246558
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 150.84,
        "prefillTokensPerSec": 2398.57,
        "decodeTokens": 6445,
        "peakRssMb": 28829,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:42:03.466Z",
        "trialId": "data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:31:41.802Z",
      "finishedAt": "2026-10-21T10:37:26.597Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 344795,
        "timing": {
          "startedAt": "2026-10-21T10:31:41.802Z",
          "finishedAt": "2026-10-21T10:37:26.597Z",
          "timeToFirstArtifactMs": 321957,
          "timeToLastArtifactWriteMs": 341536,
          "timeToFirstTokenMs": 248904,
          "firstTurnTtftMs": 10123,
          "timeToFirstToolCallMs": 251723
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 124.8,
        "prefillTokensPerSec": 2093.98,
        "decodeTokens": 8417,
        "peakRssMb": 30824,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:42:03.467Z",
        "trialId": "incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:37:36.368Z",
      "finishedAt": "2026-10-21T10:42:00.902Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 264534,
        "timing": {
          "startedAt": "2026-10-21T10:37:36.368Z",
          "finishedAt": "2026-10-21T10:42:00.902Z",
          "timeToFirstArtifactMs": 251273,
          "timeToLastArtifactWriteMs": 261720,
          "timeToFirstTokenMs": 248965,
          "firstTurnTtftMs": 10000,
          "timeToFirstToolCallMs": 249280
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 143.94,
        "prefillTokensPerSec": 2272.85,
        "decodeTokens": 1524,
        "peakRssMb": 28968,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:42:10.836Z",
        "trialId": "ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:42:10.746Z",
      "finishedAt": "2026-10-21T10:46:30.245Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 259499,
        "timing": {
          "startedAt": "2026-10-21T10:42:10.746Z",
          "finishedAt": "2026-10-21T10:46:30.245Z",
          "timeToFirstArtifactMs": 257290,
          "timeToLastArtifactWriteMs": 258999,
          "timeToFirstTokenMs": 246475,
          "firstTurnTtftMs": 7374,
          "timeToFirstToolCallMs": 246828
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 135.05,
        "prefillTokensPerSec": 2508.03,
        "decodeTokens": 1565,
        "peakRssMb": 28450,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:57:13.071Z",
        "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:46:39.157Z",
      "finishedAt": "2026-10-21T10:52:08.844Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 329687,
        "timing": {
          "startedAt": "2026-10-21T10:46:39.157Z",
          "finishedAt": "2026-10-21T10:52:08.844Z",
          "timeToFirstArtifactMs": 298402,
          "timeToLastArtifactWriteMs": 326202,
          "timeToFirstTokenMs": 248780,
          "firstTurnTtftMs": 9988,
          "timeToFirstToolCallMs": 249630
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 131.31,
        "prefillTokensPerSec": 2274.47,
        "decodeTokens": 6553,
        "peakRssMb": 31917,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:57:13.071Z",
        "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9",
        "cellId": "r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:52:18.795Z",
      "finishedAt": "2026-10-21T10:56:37.948Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 259153,
        "timing": {
          "startedAt": "2026-10-21T10:52:18.795Z",
          "finishedAt": "2026-10-21T10:56:37.948Z",
          "timeToFirstArtifactMs": 254221,
          "timeToLastArtifactWriteMs": 255466,
          "timeToFirstTokenMs": 245320,
          "firstTurnTtftMs": 6580,
          "timeToFirstToolCallMs": 246996
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 170.56,
        "prefillTokensPerSec": 2714.82,
        "decodeTokens": 1103,
        "peakRssMb": 29670,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T10:57:13.071Z",
        "trialId": "constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 1
      },
      "startedAt": "2026-10-21T10:56:47.023Z",
      "finishedAt": "2026-10-21T11:06:18.151Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 571128,
        "timing": {
          "startedAt": "2026-10-21T10:56:47.023Z",
          "finishedAt": "2026-10-21T11:06:18.151Z",
          "timeToFirstArtifactMs": 346714,
          "timeToLastArtifactWriteMs": 568868,
          "timeToFirstTokenMs": 246438,
          "firstTurnTtftMs": 7396,
          "timeToFirstToolCallMs": 246833
        },
        "toolCalls": 31,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 116.37,
        "prefillTokensPerSec": 1642.92,
        "decodeTokens": 35020,
        "peakRssMb": 31645,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:12:11.586Z",
        "trialId": "records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:06:27.563Z",
      "finishedAt": "2026-10-21T11:10:52.054Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 264491,
        "timing": {
          "startedAt": "2026-10-21T11:06:27.563Z",
          "finishedAt": "2026-10-21T11:10:52.054Z",
          "timeToFirstArtifactMs": 259324,
          "timeToLastArtifactWriteMs": 261940,
          "timeToFirstTokenMs": 246391,
          "firstTurnTtftMs": 7374,
          "timeToFirstToolCallMs": 246725
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 136.25,
        "prefillTokensPerSec": 2407.82,
        "decodeTokens": 2179,
        "peakRssMb": 28490,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:12:11.587Z",
        "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:11:00.918Z",
      "finishedAt": "2026-10-21T11:16:46.049Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 345131,
        "timing": {
          "startedAt": "2026-10-21T11:11:00.918Z",
          "finishedAt": "2026-10-21T11:16:46.049Z",
          "timeToFirstArtifactMs": 252975,
          "timeToLastArtifactWriteMs": 342188,
          "timeToFirstTokenMs": 247450,
          "firstTurnTtftMs": 8490,
          "timeToFirstToolCallMs": 248060
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 164.56,
        "prefillTokensPerSec": 2673.35,
        "decodeTokens": 1044,
        "peakRssMb": 28830,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:28:02.169Z",
        "trialId": "meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1414,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 9,
          "grounding": 8,
          "actionability": 8,
          "brevity": 9
        },
        "meanScore": 8.5,
        "justification": "Clear, compact decisions, open questions and four owned actions preserve the EU launch, batch migration and A1 dependency. The brief incorrectly says drafting Luis's copy is blocked on Legal review; the transcript only flags unscheduled review as a risk. A4's table label is ambiguous, though the open-question prose correctly says Jordan will schedule the review.",
        "artifactSha256": "91d2fba4466a31ae81eefefe933d95bcd7d196bf0bd4af99e35101c3a491ce18",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:16:55.709Z",
      "finishedAt": "2026-10-21T11:28:35.058Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 699349,
        "timing": {
          "startedAt": "2026-10-21T11:16:55.709Z",
          "finishedAt": "2026-10-21T11:28:35.058Z",
          "timeToFirstArtifactMs": 290345,
          "timeToLastArtifactWriteMs": 399517,
          "timeToFirstTokenMs": 252520,
          "firstTurnTtftMs": 8527,
          "timeToFirstToolCallMs": 254508
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 94.74,
        "prefillTokensPerSec": 1958.11,
        "decodeTokens": 38462,
        "peakRssMb": 33750,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:44:03.544Z",
        "trialId": "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 6006,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 5
        },
        "meanScore": 6.5,
        "justification": "Scoped before building; the real workflow and honest QA reached finish, but the reported scope arithmetic error stayed unfixed. Inputs update the correct base model. Scope recovery is $1,019 versus $1,110.03; every scenario repeats base ROI instead of -25.0%/-76.9%/-93.8%, and totals/legend overrun month columns.",
        "artifactSha256": "8c7ebf3c960d99e2954b39ff110315fab12653e2e600e03dd8823990207bc24b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:28:48.644Z",
      "finishedAt": "2026-10-21T11:35:44.146Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on synthesis.md; the same scenario failure stayed unchanged. Last failure: - conflict (launch budget): a superseded value appears outside the Conflicts section; re-read memo-product.md and finance.csv, then keep the historical value only in its reconciliation entry - unsupported date claim not present in the seeded sources: re-read the timing fields in memo-product.md and memo-engineering.md, then replace every launch date with a source-backed value",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 84.6% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 415502,
        "timing": {
          "startedAt": "2026-10-21T11:28:48.644Z",
          "finishedAt": "2026-10-21T11:35:44.146Z",
          "timeToFirstArtifactMs": 296705,
          "timeToLastArtifactWriteMs": 413357,
          "timeToFirstTokenMs": 248786,
          "firstTurnTtftMs": 9986,
          "timeToFirstToolCallMs": 249579
        },
        "toolCalls": 51,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 144.18,
        "prefillTokensPerSec": 1878.34,
        "decodeTokens": 14916,
        "peakRssMb": 37525,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:42:00.723Z",
        "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:35:57.371Z",
      "finishedAt": "2026-10-21T11:40:22.033Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 264662,
        "timing": {
          "startedAt": "2026-10-21T11:35:57.371Z",
          "finishedAt": "2026-10-21T11:40:22.033Z",
          "timeToFirstArtifactMs": 259430,
          "timeToLastArtifactWriteMs": 262062,
          "timeToFirstTokenMs": 247123,
          "firstTurnTtftMs": 7950,
          "timeToFirstToolCallMs": 247470
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 157.6,
        "prefillTokensPerSec": 2349.61,
        "decodeTokens": 1710,
        "peakRssMb": 30028,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:44:03.546Z",
        "trialId": "docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1225,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 8,
          "candor": 9,
          "usefulness": 8
        },
        "meanScore": 8.3,
        "justification": "Theme values, 11-to-9 pages, spacing explanation and retained Callout Emphasis match actual tool responses. The unresolved style and concrete re-keying action are clear. Repeated [1] citations misleadingly point to a placeholder explicitly containing no theme values, and the recommendation omits a final visual check.",
        "artifactSha256": "a13e33f351640c0b5a2f1061e2a11323a6e8de6d3b53f0479b9197fce7e8f372",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:40:30.846Z",
      "finishedAt": "2026-10-21T11:46:41.510Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 370664,
        "timing": {
          "startedAt": "2026-10-21T11:40:30.846Z",
          "finishedAt": "2026-10-21T11:46:41.510Z",
          "timeToFirstArtifactMs": 269898,
          "timeToLastArtifactWriteMs": 368001,
          "timeToFirstTokenMs": 253707,
          "firstTurnTtftMs": 9736,
          "timeToFirstToolCallMs": 254301
        },
        "toolCalls": 32,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 133.38,
        "prefillTokensPerSec": 2253.84,
        "decodeTokens": 9477,
        "peakRssMb": 33795,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:58:36.215Z",
        "trialId": "craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1534,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 7,
          "convertibility": 9
        },
        "meanScore": 8.5,
        "justification": "The exact brief was read and cited before outlining; all seven same-title H1 slides match the locked outline in order, and real matching PPTX copies were saved after conversion and preview. Metrics and required actions are preserved. Unseeded claims about monetary harm, expansion readiness and training effects weaken grounding.",
        "artifactSha256": "1246f1b20397774a8204a2b95ab551c99b4a319fb03719994b2eb3e5e3b3a984",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:46:51.145Z",
      "finishedAt": "2026-10-21T11:52:25.012Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 333867,
        "timing": {
          "startedAt": "2026-10-21T11:46:51.145Z",
          "finishedAt": "2026-10-21T11:52:25.012Z",
          "timeToFirstArtifactMs": 309759,
          "timeToLastArtifactWriteMs": 330327,
          "timeToFirstTokenMs": 249809,
          "firstTurnTtftMs": 6888,
          "timeToFirstToolCallMs": 261754
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 134.64,
        "prefillTokensPerSec": 2153.26,
        "decodeTokens": 7873,
        "peakRssMb": 30427,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T11:58:36.218Z",
        "trialId": "wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 6398,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 5,
          "grounding": 4,
          "citationQuality": 6,
          "calibration": 5
        },
        "meanScore": 5,
        "justification": "Chronology and three relevant source IDs are clear, and first programmer is qualified. Repetition replaces much synthesis. The cards do not support absolute claims that no program was ever executed, the memory–processor historical comparison, or the characterization of collaboration and Note G as most enduring; citations blur these additions with source facts.",
        "artifactSha256": "d229f07292e7322c6d6f953924f599ff3da39d65e949ca5ab7e464cb9d833e9e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 1
      },
      "startedAt": "2026-10-21T11:52:34.757Z",
      "finishedAt": "2026-10-21T12:20:04.478Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: themistocles/654cbfe6 exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model got stuck repeating the same action without making progress, so the turn was stopped. Try sending your message again, or rephrase your request.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 1.3,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 5,
            "ruleId": "B_CLEAN_HIGH_INTERVENTION",
            "summary": "No red flags or missing roles were observed, but the team needed at least three interventions."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1649721,
        "timing": {
          "startedAt": "2026-10-21T11:52:34.757Z",
          "finishedAt": "2026-10-21T12:20:04.478Z",
          "timeToFirstArtifactMs": 413426,
          "timeToLastArtifactWriteMs": 967216,
          "timeToFirstTokenMs": 254990,
          "firstTurnTtftMs": 11271,
          "timeToFirstToolCallMs": 256526
        },
        "toolCalls": 157,
        "autoAnswers": 10,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 89.78,
        "prefillTokensPerSec": 1537.7,
        "decodeTokens": 85966,
        "peakRssMb": 39543,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T12:29:01.535Z",
        "trialId": "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 2334,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 1,
          "prioritization": 2,
          "conflict-handling": 1,
          "realism": 1
        },
        "meanScore": 1.3,
        "justification": "The report invents Sprint Planning/Acme/Eng-lead/Roadmap events and mis-times Harborview and retro; it omits Northwind, budget/Priya and All-hands. No calendar probe ran. Its proposed clash is fictional, focus blocks overlap real meetings, and Thursday's claimed biggest gap is shorter than its own Monday/Tuesday blocks. Follow-up dates are retained.",
        "artifactSha256": "d7c5ac0667b0fbc919185497c0682d9d06c4da08abb4852a2a3d63c6a62f49ea",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 1
      },
      "startedAt": "2026-10-21T12:20:16.906Z",
      "finishedAt": "2026-10-21T12:36:14.591Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 957685,
        "timing": {
          "startedAt": "2026-10-21T12:20:16.906Z",
          "finishedAt": "2026-10-21T12:36:14.591Z",
          "timeToFirstArtifactMs": 294548,
          "timeToLastArtifactWriteMs": 956209,
          "timeToFirstTokenMs": 254510,
          "firstTurnTtftMs": 10419,
          "timeToFirstToolCallMs": 256050
        },
        "toolCalls": 100,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 97.09,
        "prefillTokensPerSec": 1702.53,
        "decodeTokens": 55464,
        "peakRssMb": 39435,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T12:42:13.725Z",
        "trialId": "craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 4245,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 1
      },
      "startedAt": "2026-10-21T12:36:26.068Z",
      "finishedAt": "2026-10-21T12:43:52.857Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 446789,
        "timing": {
          "startedAt": "2026-10-21T12:36:26.068Z",
          "finishedAt": "2026-10-21T12:43:52.857Z",
          "timeToFirstArtifactMs": 264700,
          "timeToLastArtifactWriteMs": 400800,
          "timeToFirstTokenMs": 252906,
          "firstTurnTtftMs": 8582,
          "timeToFirstToolCallMs": 254271
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 115.03,
        "prefillTokensPerSec": 2071.15,
        "decodeTokens": 17300,
        "peakRssMb": 33904,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T12:58:59.030Z",
        "trialId": "craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 7237,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 9,
          "grounding": 3,
          "structure": 7,
          "actionability": 6
        },
        "meanScore": 6.3,
        "justification": "All six entries and supplied figures are retained in APA-like form, with coherent summary/evaluation/relevance prose. Dense, unlabeled annotation paragraphs hinder scanning. Allen's single-region sample, Chen's cross-sectional design and literature claims, Dunn's usual-care comparator, and policy/technology assertions are not in the cards; several judgments rest on invented detail.",
        "artifactSha256": "ab3e79866dc092c5a33429dcc1cb51af0593ac58f1fb1bbca8f957427f893382",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 1
      },
      "startedAt": "2026-10-21T12:44:02.747Z",
      "finishedAt": "2026-10-21T12:50:13.120Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 370373,
        "timing": {
          "startedAt": "2026-10-21T12:44:02.747Z",
          "finishedAt": "2026-10-21T12:50:13.120Z",
          "timeToFirstArtifactMs": 269998,
          "timeToLastArtifactWriteMs": 322789,
          "timeToFirstTokenMs": 256498,
          "firstTurnTtftMs": 8477,
          "timeToFirstToolCallMs": 258508
        },
        "toolCalls": 40,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 129.47,
        "prefillTokensPerSec": 2172.37,
        "decodeTokens": 8299,
        "peakRssMb": 35489,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T12:58:59.032Z",
        "trialId": "craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1",
        "cellId": "r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/workspace/word-document-from-content-eval/documents/eval/report.md",
        "artifactBytes": 4442,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 2,
          "structure": 8,
          "convertibility": 8
        },
        "meanScore": 6,
        "justification": "Correct metrics, complete sections and a compact table convert to a valid saved DOCX. Grounding is weak: the report invents low cost, human responses, duplicate-refund mechanisms and causal process changes, reverses who goes silent after photos, and turns three named actions into unsupported claims about staffing and automatic review.",
        "artifactSha256": "fa1035ab92c5c6a6d9944a77a801d0db6afbb9f82c8f56f696fc1589cc38d3fa",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 1
      },
      "startedAt": "2026-10-21T12:50:26.410Z",
      "finishedAt": "2026-10-21T12:55:11.762Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 285352,
        "timing": {
          "startedAt": "2026-10-21T12:50:26.410Z",
          "finishedAt": "2026-10-21T12:55:11.762Z",
          "timeToFirstArtifactMs": 268818,
          "timeToLastArtifactWriteMs": 279170,
          "timeToFirstTokenMs": 246409,
          "firstTurnTtftMs": 7187,
          "timeToFirstToolCallMs": 247241
        },
        "toolCalls": 11,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 148.9,
        "prefillTokensPerSec": 2409.5,
        "decodeTokens": 4759,
        "peakRssMb": 28821,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T12:57:01.509Z",
        "trialId": "fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 1
      },
      "startedAt": "2026-10-21T12:55:21.219Z",
      "finishedAt": "2026-10-21T12:59:46.947Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265728,
        "timing": {
          "startedAt": "2026-10-21T12:55:21.219Z",
          "finishedAt": "2026-10-21T12:59:46.947Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 249223,
          "firstTurnTtftMs": 10336,
          "timeToFirstToolCallMs": 251519
        },
        "toolCalls": 5,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 123.35,
        "prefillTokensPerSec": 2307.76,
        "decodeTokens": 1326,
        "peakRssMb": 28898,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:12:22.235Z",
        "trialId": "dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 1
      },
      "startedAt": "2026-10-21T12:59:56.472Z",
      "finishedAt": "2026-10-21T13:09:44.963Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7.5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 588491,
        "timing": {
          "startedAt": "2026-10-21T12:59:56.472Z",
          "finishedAt": "2026-10-21T13:09:44.963Z",
          "timeToFirstArtifactMs": 260933,
          "timeToLastArtifactWriteMs": 585392,
          "timeToFirstTokenMs": 247411,
          "firstTurnTtftMs": 8453,
          "timeToFirstToolCallMs": 252021
        },
        "toolCalls": 23,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 137.68,
        "prefillTokensPerSec": 2290.23,
        "decodeTokens": 10007,
        "peakRssMb": 33518,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:12:22.235Z",
        "trialId": "interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 1
      },
      "startedAt": "2026-10-21T13:09:58.116Z",
      "finishedAt": "2026-10-21T13:15:59.512Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 361396,
        "timing": {
          "startedAt": "2026-10-21T13:09:58.116Z",
          "finishedAt": "2026-10-21T13:15:59.512Z",
          "timeToFirstArtifactMs": 333881,
          "timeToLastArtifactWriteMs": 333881,
          "timeToFirstTokenMs": 252298,
          "firstTurnTtftMs": 8303,
          "timeToFirstToolCallMs": 252840
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 127.96,
        "prefillTokensPerSec": 2263.29,
        "decodeTokens": 9594,
        "peakRssMb": 32116,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:28:28.978Z",
        "trialId": "craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 3279,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 8,
          "usefulness": 9
        },
        "meanScore": 8.5,
        "justification": "The snapshot-based scope/report/verdict workflow is complete. The concise, ranked report catches both SQL regressions, logging, the leaked handle, false-success catch and pagination bound, with concrete fixes; all three deliberate patterns are correctly accepted. The explicitly required payment.js workspace read occurred only after the report and verdict.",
        "artifactSha256": "cdab9bea0ed8e1a5e6097fbcb49c2762509e14728dd79187b7e3a2250482d665",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 1
      },
      "startedAt": "2026-10-21T13:16:09.239Z",
      "finishedAt": "2026-10-21T13:24:29.221Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 499982,
        "timing": {
          "startedAt": "2026-10-21T13:16:09.239Z",
          "finishedAt": "2026-10-21T13:24:29.221Z",
          "timeToFirstArtifactMs": 318750,
          "timeToLastArtifactWriteMs": 436995,
          "timeToFirstTokenMs": 256257,
          "firstTurnTtftMs": 8769,
          "timeToFirstToolCallMs": 257513
        },
        "toolCalls": 32,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 105.32,
        "prefillTokensPerSec": 1969.61,
        "decodeTokens": 19220,
        "peakRssMb": 34913,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:28:28.980Z",
        "trialId": "craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 10096,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 9,
          "usefulness": 6
        },
        "meanScore": 7.5,
        "justification": "Required inputs precede the scoped review/report workflow. It finds missing history handling, numeric retryCount and undocumented widget GET, while accepting sessions and optional nullable email. Compatibility is overstated: 1.0.0 does not prove a new API; field renaming, type/status changes and a newly required input contradict the no-breaking-changes verdict.",
        "artifactSha256": "e8e79b11fa7ded204357c68c45f2a61479a573ef2f7856bedd31768721b086a6",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 1
      },
      "startedAt": "2026-10-21T13:24:40.873Z",
      "finishedAt": "2026-10-21T13:30:32.541Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-deep-security-review passed 11 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 351668,
        "timing": {
          "startedAt": "2026-10-21T13:24:40.873Z",
          "finishedAt": "2026-10-21T13:30:32.541Z",
          "timeToFirstArtifactMs": 268213,
          "timeToLastArtifactWriteMs": 347841,
          "timeToFirstTokenMs": 248398,
          "firstTurnTtftMs": 9054,
          "timeToFirstToolCallMs": 249453
        },
        "toolCalls": 26,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 138.24,
        "prefillTokensPerSec": 2011.74,
        "decodeTokens": 8146,
        "peakRssMb": 33343,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:37:58.994Z",
        "trialId": "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 6346,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 6,
          "remediation": 8
        },
        "meanScore": 7.3,
        "justification": "Static paths confirm all three admin defects; safe profile/reports constructs are correctly excluded. Fixes use admin middleware, bound SQL and shell-free argument arrays. Blast-radius claims assume deployment reachability/DB powers, CVSS numbers lack vectors, and the narrative calls SQL critical while the table/JSON rate it high. Runtime exploitation was correctly left unclaimed.",
        "artifactSha256": "7b4fc860bf7217843a54bf166bd3634bdafed7f4a259a1d533e2d67702e0c59d",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 1
      },
      "startedAt": "2026-10-21T13:30:42.019Z",
      "finishedAt": "2026-10-21T13:38:23.311Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 461292,
        "timing": {
          "startedAt": "2026-10-21T13:30:42.019Z",
          "finishedAt": "2026-10-21T13:38:23.311Z",
          "timeToFirstArtifactMs": 374938,
          "timeToLastArtifactWriteMs": 455960,
          "timeToFirstTokenMs": 254443,
          "firstTurnTtftMs": 11241,
          "timeToFirstToolCallMs": 254892
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 82.24,
        "prefillTokensPerSec": 1140.39,
        "decodeTokens": 11516,
        "peakRssMb": 33553,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:39:12.305Z",
        "trialId": "large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 1
      },
      "startedAt": "2026-10-21T13:38:35.294Z",
      "finishedAt": "2026-10-21T13:44:55.177Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 379883,
        "timing": {
          "startedAt": "2026-10-21T13:38:35.294Z",
          "finishedAt": "2026-10-21T13:44:55.177Z",
          "timeToFirstArtifactMs": 273380,
          "timeToLastArtifactWriteMs": 374820,
          "timeToFirstTokenMs": 253103,
          "firstTurnTtftMs": 8746,
          "timeToFirstToolCallMs": 255237
        },
        "toolCalls": 66,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 134.16,
        "prefillTokensPerSec": 2141.5,
        "decodeTokens": 10275,
        "peakRssMb": 35433,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T13:58:40.961Z",
        "trialId": "craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 2977,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 9,
          "Invariant discipline": 9,
          "Reconciled verification": 9
        },
        "meanScore": 9,
        "justification": "Real pre-edit searches found all 11 mentions: seven rename sites plus four documented historical/spec exclusions. The four changed files match the exact mechanical rename and exclusions stay unchanged. A real residual source/test scan reports zero matches; 7+4=11 reconciles, and three captured suite runs pass 3/3, including two after edits.",
        "artifactSha256": "4403f866a3ae99fedb70195593b57a3dc08b2711c5799a01fb46c13ff5f4d0e1",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 1
      },
      "startedAt": "2026-10-21T13:45:07.702Z",
      "finishedAt": "2026-10-21T14:05:59.289Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1251587,
        "timing": {
          "startedAt": "2026-10-21T13:45:07.702Z",
          "finishedAt": "2026-10-21T14:05:59.289Z",
          "timeToFirstArtifactMs": 273580,
          "timeToLastArtifactWriteMs": 1230600,
          "timeToFirstTokenMs": 268636,
          "firstTurnTtftMs": 7534,
          "timeToFirstToolCallMs": 271120
        },
        "toolCalls": 92,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.57,
        "prefillTokensPerSec": 1341.42,
        "decodeTokens": 61297,
        "peakRssMb": 35097,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T14:18:35.721Z",
        "trialId": "craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 1990,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 8,
          "Reversible-stage execution": 5,
          "Genuine deduplication": 9
        },
        "meanScore": 7.3,
        "justification": "Real 13-test green baselines and stated 2×6-line counts precede production edits; final shared helper preserves behavior and all exports. No test ran between wiring the two callers; a duplicate import was repaired before testing. Per-stage green claims and “no deviations” overstate the captured execution.",
        "artifactSha256": "2176e57a585796f5bb173eeda9ca69b8b5971a09962a5180264cc811669a5844",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 1
      },
      "startedAt": "2026-10-21T14:06:11.935Z",
      "finishedAt": "2026-10-21T14:30:30.724Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "chat stalled for 305s — daemon reachable but issuing no model turns (soft digest stuck at a8c34f780ff2a339) (re-engage nudge was sent and ignored)",
        "failureMode": "chat-stalled",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.6,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 96.4% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1458789,
        "timing": {
          "startedAt": "2026-10-21T14:06:11.935Z",
          "finishedAt": "2026-10-21T14:30:30.724Z",
          "timeToFirstArtifactMs": 279076,
          "timeToLastArtifactWriteMs": 959637,
          "timeToFirstTokenMs": 263865,
          "firstTurnTtftMs": 9254,
          "timeToFirstToolCallMs": 264886
        },
        "toolCalls": 66,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 60.75,
        "prefillTokensPerSec": 1271.56,
        "decodeTokens": 22254,
        "peakRssMb": 35180,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T14:56:01.245Z",
        "trialId": "craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk",
        "cellId": "r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 2754,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 8,
          "Honest verification": 6
        },
        "meanScore": 7.7,
        "justification": "Captured test fails 80≠68 before the one-character >= fix, then all four tests pass; original tests survive. Package-script churn removes the seeded final newline, contradicting “no change” and causing the fixed gate failure. Review misses that mismatch; validation’s “nothing unverified” conflicts with the disclosed missing mutant run.",
        "artifactSha256": "9529f80e7c5265e0637479106ee532e26915c879a983269b1bbab4f3031b04ba",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 1
      },
      "startedAt": "2026-10-21T14:30:43.336Z",
      "finishedAt": "2026-10-21T14:35:07.024Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 263688,
        "timing": {
          "startedAt": "2026-10-21T14:30:43.336Z",
          "finishedAt": "2026-10-21T14:35:07.024Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 250856,
          "firstTurnTtftMs": 10435,
          "timeToFirstToolCallMs": 251926
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 126.31,
        "prefillTokensPerSec": 2539.19,
        "decodeTokens": 825,
        "peakRssMb": 28866,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T14:53:38.817Z",
        "trialId": "craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 1
      },
      "startedAt": "2026-10-21T14:35:18.056Z",
      "finishedAt": "2026-10-21T14:56:32.001Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1273945,
        "timing": {
          "startedAt": "2026-10-21T14:35:18.056Z",
          "finishedAt": "2026-10-21T14:56:32.001Z",
          "timeToFirstArtifactMs": 937683,
          "timeToLastArtifactWriteMs": 1251400,
          "timeToFirstTokenMs": 252528,
          "firstTurnTtftMs": 12931,
          "timeToFirstToolCallMs": 253829
        },
        "toolCalls": 59,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 68.36,
        "prefillTokensPerSec": 1063.36,
        "decodeTokens": 53625,
        "peakRssMb": 36830,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T14:56:54.226Z",
        "trialId": "craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 1
      },
      "startedAt": "2026-10-21T14:56:44.141Z",
      "finishedAt": "2026-10-21T15:04:44.126Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 479985,
        "timing": {
          "startedAt": "2026-10-21T14:56:44.141Z",
          "finishedAt": "2026-10-21T15:04:44.126Z",
          "timeToFirstArtifactMs": 295670,
          "timeToLastArtifactWriteMs": 473477,
          "timeToFirstTokenMs": 257445,
          "firstTurnTtftMs": 8768,
          "timeToFirstToolCallMs": 258482
        },
        "toolCalls": 46,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 123.28,
        "prefillTokensPerSec": 1748.8,
        "decodeTokens": 18114,
        "peakRssMb": 37090,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T15:16:01.320Z",
        "trialId": "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 711,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 9,
          "numbering continuity": 9,
          "professionalism": 2,
          "completeness": 4,
          "print-polish": 8
        },
        "meanScore": 6.4,
        "justification": "Ledger amounts total 5,415 and numbers 042–044 are correct; all three print cleanly on single Letter pages. Nov 1 issue dates conflict with due dates/19-day terms; contacts and the 044 header are wrong. Books remain at 041 with all three entries uninvoiced; the report’s “went out” claim is unsupported.",
        "artifactSha256": "3ca039c1c7f77e1f5474ff95f9910e7b2f4736655c3830336f59e7d2fb94a6f3",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call."
      }
    },
    {
      "trialId": "craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 1
      },
      "startedAt": "2026-10-21T15:04:53.129Z",
      "finishedAt": "2026-10-21T15:33:23.016Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1709887,
        "timing": {
          "startedAt": "2026-10-21T15:04:53.129Z",
          "finishedAt": "2026-10-21T15:33:23.016Z",
          "timeToFirstArtifactMs": 985927,
          "timeToLastArtifactWriteMs": 1690501,
          "timeToFirstTokenMs": 257373,
          "firstTurnTtftMs": 13156,
          "timeToFirstToolCallMs": 258636
        },
        "toolCalls": 89,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 66.84,
        "prefillTokensPerSec": 1116.82,
        "decodeTokens": 47786,
        "peakRssMb": 38720,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T15:43:05.592Z",
        "trialId": "craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 1
      },
      "startedAt": "2026-10-21T15:33:33.598Z",
      "finishedAt": "2026-10-21T15:40:04.944Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 391346,
        "timing": {
          "startedAt": "2026-10-21T15:33:33.598Z",
          "finishedAt": "2026-10-21T15:40:04.944Z",
          "timeToFirstArtifactMs": 288230,
          "timeToLastArtifactWriteMs": 315198,
          "timeToFirstTokenMs": 250440,
          "firstTurnTtftMs": 10354,
          "timeToFirstToolCallMs": 255109
        },
        "toolCalls": 44,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 132.12,
        "prefillTokensPerSec": 2189.67,
        "decodeTokens": 11962,
        "peakRssMb": 37970,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T15:43:05.593Z",
        "trialId": "craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 1
      },
      "startedAt": "2026-10-21T15:40:14.343Z",
      "finishedAt": "2026-10-21T16:06:11.724Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on craftbook: supplier intake recipe; missing signals stayed unchanged: task supplier-intake/1 (from craftbook \"supplier-intake\") has status \"paused\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.. Last failure: task supplier-intake/1 (from craftbook \"supplier-intake\") has status \"paused\" — advance THIS task to a terminal step with advance_task_step/set_task_status. Do NOT invoke the recipe again or create a second task: a new task starts the work over and does not count.",
        "failureMode": "model-stuck",
        "failureClass": "infra",
        "includedInFixedModelAggregate": false,
        "failureClassRule": "gate-unevaluable"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.3,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 83.3% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1557381,
        "timing": {
          "startedAt": "2026-10-21T15:40:14.343Z",
          "finishedAt": "2026-10-21T16:06:11.724Z",
          "timeToFirstArtifactMs": 637191,
          "timeToLastArtifactWriteMs": 1555423,
          "timeToFirstTokenMs": 256578,
          "firstTurnTtftMs": 13196,
          "timeToFirstToolCallMs": 257230
        },
        "toolCalls": 84,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 70.97,
        "prefillTokensPerSec": 948.36,
        "decodeTokens": 42887,
        "peakRssMb": 38359,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T16:17:13.052Z",
        "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 1
      },
      "startedAt": "2026-10-21T16:18:23.023Z",
      "finishedAt": "2026-10-21T17:24:46.607Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3983584,
        "timing": {
          "startedAt": "2026-10-21T16:18:23.023Z",
          "finishedAt": "2026-10-21T17:24:46.607Z",
          "timeToFirstArtifactMs": 422112,
          "timeToLastArtifactWriteMs": 3979042,
          "timeToFirstTokenMs": 253629,
          "firstTurnTtftMs": 13167,
          "timeToFirstToolCallMs": 254666
        },
        "toolCalls": 116,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 64.64,
        "prefillTokensPerSec": 983.24,
        "decodeTokens": 96407,
        "peakRssMb": 34511,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T17:28:01.187Z",
        "trialId": "craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 1
      },
      "startedAt": "2026-10-21T17:24:59.702Z",
      "finishedAt": "2026-10-21T18:29:08.867Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-gate-script:4:targetnone:fr600bdd:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 11 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "infra",
        "includedInFixedModelAggregate": false,
        "failureClassRule": "gate-unevaluable"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.7,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3849165,
        "timing": {
          "startedAt": "2026-10-21T17:24:59.702Z",
          "finishedAt": "2026-10-21T18:29:08.867Z",
          "timeToFirstArtifactMs": 2162735,
          "timeToLastArtifactWriteMs": 2760534,
          "timeToFirstTokenMs": 252898,
          "firstTurnTtftMs": 13041,
          "timeToFirstToolCallMs": 253982
        },
        "toolCalls": 78,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 37.88,
        "prefillTokensPerSec": 612.13,
        "decodeTokens": 81173,
        "peakRssMb": 39229,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T18:46:45.459Z",
        "trialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb",
      "cell": {
        "id": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-fanout",
        "repetition": 1
      },
      "startedAt": "2026-10-21T18:47:26.961Z",
      "finishedAt": "2026-10-21T19:41:37.095Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-fanout:6:targetnone:fr1tjwffw:rp0:rf0:m2\" stuck for 18m — artifact scored but stalled 18m despite 16 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.6,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 75% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 3250134,
        "timing": {
          "startedAt": "2026-10-21T18:47:26.961Z",
          "finishedAt": "2026-10-21T19:41:37.095Z",
          "timeToFirstArtifactMs": 718102,
          "timeToLastArtifactWriteMs": 1503103,
          "timeToFirstTokenMs": 255059,
          "firstTurnTtftMs": 13811,
          "timeToFirstToolCallMs": 255871
        },
        "toolCalls": 89,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 55.59,
        "prefillTokensPerSec": 814.58,
        "decodeTokens": 46890,
        "peakRssMb": 39260,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T19:43:07.197Z",
        "trialId": "craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb",
        "cellId": "r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout",
        "scenarioId": "craftbook-author-fanout",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-tictactoe",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "tictactoe",
        "repetition": 2
      },
      "startedAt": "2026-10-21T19:41:53.524Z",
      "finishedAt": "2026-10-21T19:46:27.602Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 274078,
        "timing": {
          "startedAt": "2026-10-21T19:41:53.524Z",
          "finishedAt": "2026-10-21T19:46:27.602Z",
          "timeToFirstArtifactMs": 266293,
          "timeToLastArtifactWriteMs": 266297,
          "timeToFirstTokenMs": 250070,
          "firstTurnTtftMs": 8849,
          "timeToFirstToolCallMs": 252082
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 176.38,
        "prefillTokensPerSec": 2764.38,
        "decodeTokens": 1983,
        "peakRssMb": 29079,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:03:29.904Z",
        "trialId": "tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-tictactoe",
        "scenarioId": "tictactoe",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/workspace/tic-tac-toe-game/index.html",
        "artifactBytes": 4466,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 6,
          "functionalCompleteness": 9,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.8,
        "justification": "All eight X lines, O win, draw, occupied-cell protection, terminal locking and reset work. The compact board fits mobile; color-coded marks and win highlighting are clear. Styling is basic, empty cells have no accessible names, and turn/winner status has no live announcements.",
        "artifactSha256": "f60468fea3daa6fae9f3922a35627da3156aa7be5c2e5d6eb5c20fff88faa88e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tictactoe-review-ornith-r2-2026-10-21T1957.json"
      }
    },
    {
      "trialId": "petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-petshop",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "petshop",
        "repetition": 2
      },
      "startedAt": "2026-10-21T19:47:03.112Z",
      "finishedAt": "2026-10-21T19:52:31.157Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 328045,
        "timing": {
          "startedAt": "2026-10-21T19:47:03.112Z",
          "finishedAt": "2026-10-21T19:52:31.157Z",
          "timeToFirstArtifactMs": 323639,
          "timeToLastArtifactWriteMs": 323639,
          "timeToFirstTokenMs": 248953,
          "firstTurnTtftMs": 8879,
          "timeToFirstToolCallMs": 278360
        },
        "toolCalls": 8,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 137.73,
        "prefillTokensPerSec": 2627.68,
        "decodeTokens": 8391,
        "peakRssMb": 36703,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:03:29.906Z",
        "trialId": "petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-petshop",
        "scenarioId": "petshop",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/workspace/pet-shop-website/index.html",
        "artifactBytes": 11326,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 7,
          "functionalCompleteness": 7,
          "codeQuality": 7,
          "polish": 5
        },
        "meanScore": 6.5,
        "justification": "A readable, responsive page with six product cards, three services and the generated local PNG. Capture shows logo generation before the only HTML write, violating the explicit order. Emoji art and placeholder business claims feel generic; mobile header is flush left, and sticky navigation lacks anchor offset.",
        "artifactSha256": "6afec9afd8180c72211043ab9d970ce8589e09fd0551de6f9f431af397b46fa8",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/petshop-review-ornith-r2-2026-10-21T1957.json"
      }
    },
    {
      "trialId": "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-tankcombat",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "tankcombat",
        "repetition": 2
      },
      "startedAt": "2026-10-21T19:53:07.040Z",
      "finishedAt": "2026-10-21T19:57:54.457Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "tank-combat-arcade/workspace/index.html passed sniff (signals: tank-vocab, render-surface, keyboard-input, combat, game-loop, gameplay, js-parses, js-size-ok, html-size-ok) + runtime (3 assertion(s))",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 287417,
        "timing": {
          "startedAt": "2026-10-21T19:53:07.040Z",
          "finishedAt": "2026-10-21T19:57:54.457Z",
          "timeToFirstArtifactMs": 279900,
          "timeToLastArtifactWriteMs": 279904,
          "timeToFirstTokenMs": 252585,
          "firstTurnTtftMs": 8976,
          "timeToFirstToolCallMs": 255921
        },
        "toolCalls": 2,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 159.43,
        "prefillTokensPerSec": 2753.34,
        "decodeTokens": 3881,
        "peakRssMb": 29453,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:03:29.906Z",
        "trialId": "tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-tankcombat",
        "scenarioId": "tankcombat",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/workspace/tank-combat-arcade/index.html",
        "artifactBytes": 8633,
        "artifactLimitBytes": 20480,
        "rubric": "default-html",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "visualQuality": 5,
          "functionalCompleteness": 2,
          "codeQuality": 2,
          "polish": 3
        },
        "meanScore": 3,
        "justification": "Movement, enemy tanks, HUD and restart render, but collision code is never called and score can never increase. Repeated bullet y-updates cancel player shots to zero motion; real-time replay confirms every shot stays at y=503. Firing cannot stop enemies; seeded replay ends at zero score. Fixed canvas clips on mobile.",
        "artifactSha256": "e155b1515c49565afdb9def9bd4328ac4ede5ae0b567ceac3732f3ae1c8e8b95",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tankcombat-review-ornith-r2-2026-10-21T1957.json"
      }
    },
    {
      "trialId": "schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-schema-migration",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "schema-migration",
        "repetition": 2
      },
      "startedAt": "2026-10-21T19:58:30.243Z",
      "finishedAt": "2026-10-21T20:04:07.814Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 337571,
        "timing": {
          "startedAt": "2026-10-21T19:58:30.243Z",
          "finishedAt": "2026-10-21T20:04:07.814Z",
          "timeToFirstArtifactMs": 248401,
          "timeToLastArtifactWriteMs": 333381,
          "timeToFirstTokenMs": 247431,
          "firstTurnTtftMs": 7170,
          "timeToFirstToolCallMs": 248403
        },
        "toolCalls": 21,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 158.25,
        "prefillTokensPerSec": 2258.75,
        "decodeTokens": 3602,
        "peakRssMb": 33746,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:13:03.712Z",
        "trialId": "schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-schema-migration",
        "scenarioId": "schema-migration",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "failing-tests-spec",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:04:18.037Z",
      "finishedAt": "2026-10-21T20:12:46.503Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 508466,
        "timing": {
          "startedAt": "2026-10-21T20:04:18.037Z",
          "finishedAt": "2026-10-21T20:12:46.503Z",
          "timeToFirstArtifactMs": 253223,
          "timeToLastArtifactWriteMs": 502669,
          "timeToFirstTokenMs": 247823,
          "firstTurnTtftMs": 7668,
          "timeToFirstToolCallMs": 248484
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 170.97,
        "prefillTokensPerSec": 2728.87,
        "decodeTokens": 1127,
        "peakRssMb": 29557,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:13:03.713Z",
        "trialId": "failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec",
        "scenarioId": "failing-tests-spec",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-symptom-debug",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "symptom-debug",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:12:56.468Z",
      "finishedAt": "2026-10-21T20:17:41.485Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 285017,
        "timing": {
          "startedAt": "2026-10-21T20:12:56.468Z",
          "finishedAt": "2026-10-21T20:17:41.485Z",
          "timeToFirstArtifactMs": 280519,
          "timeToLastArtifactWriteMs": 280519,
          "timeToFirstTokenMs": 252142,
          "firstTurnTtftMs": 7808,
          "timeToFirstToolCallMs": 252651
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 143.82,
        "prefillTokensPerSec": 2698.99,
        "decodeTokens": 3997,
        "peakRssMb": 28359,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:27:56.740Z",
        "trialId": "symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-symptom-debug",
        "scenarioId": "symptom-debug",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-data-wrangle",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "data-wrangle",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:17:51.556Z",
      "finishedAt": "2026-10-21T20:22:32.290Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "out/customers.json passed all 11 property checks (parses-as-array, row-shape, emails-normalized, dates-iso, unique-emails, source-ids-preserved, row-count, sorted-by-email, golden-bom-row, golden-quoted-name, golden-dedup-newest)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 280734,
        "timing": {
          "startedAt": "2026-10-21T20:17:51.556Z",
          "finishedAt": "2026-10-21T20:22:32.290Z",
          "timeToFirstArtifactMs": 272837,
          "timeToLastArtifactWriteMs": 272837,
          "timeToFirstTokenMs": 248019,
          "firstTurnTtftMs": 7840,
          "timeToFirstToolCallMs": 249300
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 149.95,
        "prefillTokensPerSec": 2542.43,
        "decodeTokens": 4172,
        "peakRssMb": 28584,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:27:56.741Z",
        "trialId": "data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-data-wrangle",
        "scenarioId": "data-wrangle",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-incident-postmortem",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "incident-postmortem",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:22:41.378Z",
      "finishedAt": "2026-10-21T20:27:42.745Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all hard signals firing (8/8; advisory no-blame-language=pass; signals: file-present, all-sections, summary-concise, evidence-citations, timestamp-citations, grounded-core-facts, no-unsupported-certainty, action-items-formatted, no-blame-language)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 301367,
        "timing": {
          "startedAt": "2026-10-21T20:22:41.378Z",
          "finishedAt": "2026-10-21T20:27:42.745Z",
          "timeToFirstArtifactMs": 297813,
          "timeToLastArtifactWriteMs": 297813,
          "timeToFirstTokenMs": 250792,
          "firstTurnTtftMs": 10639,
          "timeToFirstToolCallMs": 252299
        },
        "toolCalls": 10,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 108.58,
        "prefillTokensPerSec": 2181.45,
        "decodeTokens": 4730,
        "peakRssMb": 28865,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:27:56.741Z",
        "trialId": "incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-incident-postmortem",
        "scenarioId": "incident-postmortem",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "ops-runbook-anomaly",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:27:52.771Z",
      "finishedAt": "2026-10-21T20:32:33.748Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "runbook executed with verbatim records and a correct halt on the step-4 anomaly (signals: stopped-on-anomaly, step1-recorded, step2-recorded, step3-recorded, no-phantom-completion, halt-report-grounded, source-reads-grounded)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 280977,
        "timing": {
          "startedAt": "2026-10-21T20:27:52.771Z",
          "finishedAt": "2026-10-21T20:32:33.748Z",
          "timeToFirstArtifactMs": 253877,
          "timeToLastArtifactWriteMs": 276984,
          "timeToFirstTokenMs": 251230,
          "firstTurnTtftMs": 10788,
          "timeToFirstToolCallMs": 251576
        },
        "toolCalls": 19,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 131.53,
        "prefillTokensPerSec": 2054.95,
        "decodeTokens": 2029,
        "peakRssMb": 30650,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:42:57.324Z",
        "trialId": "ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly",
        "scenarioId": "ops-runbook-anomaly",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:32:50.178Z",
      "finishedAt": "2026-10-21T20:37:15.917Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265739,
        "timing": {
          "startedAt": "2026-10-21T20:32:50.178Z",
          "finishedAt": "2026-10-21T20:37:15.917Z",
          "timeToFirstArtifactMs": 261154,
          "timeToLastArtifactWriteMs": 261154,
          "timeToFirstTokenMs": 248574,
          "firstTurnTtftMs": 8413,
          "timeToFirstToolCallMs": 248881
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 113.92,
        "prefillTokensPerSec": 2316.98,
        "decodeTokens": 1558,
        "peakRssMb": 28466,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:42:57.325Z",
        "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "core",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:37:27.734Z",
      "finishedAt": "2026-10-21T20:42:59.006Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: tamsin/555136ed exhausted its bounded automatic recovery allowance at the current checked progress; last error: The model's `read_file` calls failed 5 times in a row, so the turn was stopped. Try sending your message again.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "tool-repeat-abort-storm"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 6,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 9.2,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 92.3% of its measured signals."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 331272,
        "timing": {
          "startedAt": "2026-10-21T20:37:27.734Z",
          "finishedAt": "2026-10-21T20:42:59.006Z",
          "timeToFirstArtifactMs": 295675,
          "timeToLastArtifactWriteMs": 320331,
          "timeToFirstTokenMs": 250810,
          "firstTurnTtftMs": 10489,
          "timeToFirstToolCallMs": 251684
        },
        "toolCalls": 25,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 138.02,
        "prefillTokensPerSec": 2213.93,
        "decodeTokens": 6386,
        "peakRssMb": 33310,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:43:32.048Z",
        "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2",
        "cellId": "r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "core",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "constrained-comms",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:43:09.029Z",
      "finishedAt": "2026-10-21T20:47:29.727Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "notice satisfies the full constraint stack (signals: word-band, payment-data-disclosure, support-contact, service-credit, no-guarantee, no-never-again, no-100-secure, no-unhackable, grounded-facts)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 260698,
        "timing": {
          "startedAt": "2026-10-21T20:43:09.029Z",
          "finishedAt": "2026-10-21T20:47:29.727Z",
          "timeToFirstArtifactMs": 257311,
          "timeToLastArtifactWriteMs": 259189,
          "timeToFirstTokenMs": 247308,
          "firstTurnTtftMs": 7061,
          "timeToFirstToolCallMs": 249206
        },
        "toolCalls": 7,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 163.23,
        "prefillTokensPerSec": 2609.79,
        "decodeTokens": 1300,
        "peakRssMb": 29560,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:58:01.469Z",
        "trialId": "constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms",
        "scenarioId": "constrained-comms",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-records-intake",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "records-intake",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:47:38.794Z",
      "finishedAt": "2026-10-21T20:55:30.784Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "attendees.csv matches the declared schema with all golden rows (signals: csv-shape, record-schema, golden-rows)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 471990,
        "timing": {
          "startedAt": "2026-10-21T20:47:38.794Z",
          "finishedAt": "2026-10-21T20:55:30.784Z",
          "timeToFirstArtifactMs": 312052,
          "timeToLastArtifactWriteMs": 469525,
          "timeToFirstTokenMs": 247910,
          "firstTurnTtftMs": 7909,
          "timeToFirstToolCallMs": 248568
        },
        "toolCalls": 30,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 116.82,
        "prefillTokensPerSec": 1663.29,
        "decodeTokens": 23485,
        "peakRssMb": 32415,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T20:58:01.470Z",
        "trialId": "records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-records-intake",
        "scenarioId": "records-intake",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "plan-and-estimate",
        "repetition": 2
      },
      "startedAt": "2026-10-21T20:55:40.600Z",
      "finishedAt": "2026-10-21T21:00:06.569Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265969,
        "timing": {
          "startedAt": "2026-10-21T20:55:40.600Z",
          "finishedAt": "2026-10-21T21:00:06.569Z",
          "timeToFirstArtifactMs": 264465,
          "timeToLastArtifactWriteMs": 264465,
          "timeToFirstTokenMs": 249135,
          "firstTurnTtftMs": 8407,
          "timeToFirstToolCallMs": 249391
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 112.04,
        "prefillTokensPerSec": 2606.3,
        "decodeTokens": 1678,
        "peakRssMb": 28354,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T21:12:57.589Z",
        "trialId": "plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate",
        "scenarioId": "plan-and-estimate",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "meeting-followup",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:00:21.980Z",
      "finishedAt": "2026-10-21T21:06:22.948Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "meeting brief and action register pass all reconciliation gates (ordered-sections, launch-decision, migration-decision, annual-plan-open, rollback-owner-open, legal-risk, stale-proposals-rejected, csv-header, four-actions, action-row-a1, action-row-a2, action-row-a3, action-row-a4)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 360968,
        "timing": {
          "startedAt": "2026-10-21T21:00:21.980Z",
          "finishedAt": "2026-10-21T21:06:22.948Z",
          "timeToFirstArtifactMs": 255992,
          "timeToLastArtifactWriteMs": 356897,
          "timeToFirstTokenMs": 249451,
          "firstTurnTtftMs": 9624,
          "timeToFirstToolCallMs": 249960
        },
        "toolCalls": 6,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 145.7,
        "prefillTokensPerSec": 2450.08,
        "decodeTokens": 1368,
        "peakRssMb": 29822,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T21:14:08.250Z",
        "trialId": "meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup",
        "scenarioId": "meeting-followup",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/workspace/orchid-launch-follow-up/meeting-brief.md",
        "artifactBytes": 1614,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "decisionClarity": 9,
          "grounding": 8,
          "actionability": 8,
          "brevity": 9
        },
        "meanScore": 8.5,
        "justification": "Three launch decisions, open annual-plan/rollback questions and Legal risk are clearly separated. Owners, dates and A2→A1 dependency match the transcript; stale proposals stay rejected. The compact brief and CSV are usable, but A4 names a decision review rather than scheduling it by Aug 30, despite the open-question clarification.",
        "artifactSha256": "336a4c0c85483473bfff37ed0cac1fa582043aa356e32a8d62c2e3174ae80051",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/meeting-review-ornith-r2-2026-10-21T2112.json"
      }
    },
    {
      "trialId": "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-spreadsheet-model",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:06:34.092Z",
      "finishedAt": "2026-10-21T21:32:05.708Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-spreadsheet-model passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1531616,
        "timing": {
          "startedAt": "2026-10-21T21:06:34.092Z",
          "finishedAt": "2026-10-21T21:32:05.708Z",
          "timeToFirstArtifactMs": 310692,
          "timeToLastArtifactWriteMs": 548976,
          "timeToFirstTokenMs": 257590,
          "firstTurnTtftMs": 10075,
          "timeToFirstToolCallMs": 259817
        },
        "toolCalls": 49,
        "autoAnswers": 3,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 83.04,
        "prefillTokensPerSec": 1472.23,
        "decodeTokens": 95777,
        "peakRssMb": 37438,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T21:45:35.236Z",
        "trialId": "craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model",
        "scenarioId": "craftbook-spreadsheet-model",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/artifacts/spreadsheet-financial-model-workflow-eval/tasks/1/scope.md",
        "artifactBytes": 5888,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 3
        },
        "meanScore": 5,
        "justification": "The named scope→build→evaluate loop ran and live inputs recompute, but the saved contract still claims $82,752/8.77mo versus the actual $33,001.36/21.82mo. QA identified this yet finished without the fix. Revenue 70 is falsely attributed to a CSV mean of 1190; response delay is monetized as labor, and recurring staffing is reused as payback capital.",
        "artifactSha256": "209262a3ae1be8466c239677c59989788f05d07ee0b349af5f6ce93c36f92243",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/spreadsheet-review-ornith-r2-2026-10-21T2142.json"
      }
    },
    {
      "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "conflict-synthesis",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:32:19.179Z",
      "finishedAt": "2026-10-21T21:39:28.315Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all three conflicts reconciled explicitly with citations (signals: ordered-sections, conflict-launch-date-surfaced, conflict-launch-date-quarantined, conflict-launch-date-canonical-used, conflict-budget-surfaced, conflict-budget-quarantined, conflict-budget-canonical-used, conflict-dri-surfaced, conflict-dri-quarantined, conflict-dri-canonical-used, claims-source-supported, source-reads-grounded, citations)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 429136,
        "timing": {
          "startedAt": "2026-10-21T21:32:19.179Z",
          "finishedAt": "2026-10-21T21:39:28.315Z",
          "timeToFirstArtifactMs": 311912,
          "timeToLastArtifactWriteMs": 424232,
          "timeToFirstTokenMs": 252304,
          "firstTurnTtftMs": 12081,
          "timeToFirstToolCallMs": 253972
        },
        "toolCalls": 28,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 77.26,
        "prefillTokensPerSec": 1517.55,
        "decodeTokens": 10749,
        "peakRssMb": 31950,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T21:43:04.030Z",
        "trialId": "conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis",
        "scenarioId": "conflict-synthesis",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "docblocks-theme-roundtrip",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:40:00.539Z",
      "finishedAt": "2026-10-21T21:45:01.449Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "theme report is tool-grounded and the themed deck is a real PPTX (ordered-sections, theme-id, fonts, palette, page-delta, style-conflict, conflict-not-whitewashed, recommendation, docblocks-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 300910,
        "timing": {
          "startedAt": "2026-10-21T21:40:00.539Z",
          "finishedAt": "2026-10-21T21:45:01.449Z",
          "timeToFirstArtifactMs": 296178,
          "timeToLastArtifactWriteMs": 296178,
          "timeToFirstTokenMs": 260005,
          "firstTurnTtftMs": 16790,
          "timeToFirstToolCallMs": 261204
        },
        "toolCalls": 13,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 61.37,
        "prefillTokensPerSec": 694.52,
        "decodeTokens": 1533,
        "peakRssMb": 30743,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T21:46:38.204Z",
        "trialId": "docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip",
        "scenarioId": "docblocks-theme-roundtrip",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/workspace/brand-theme-alignment/theme-report.md",
        "artifactBytes": 1430,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 8,
          "candor": 9,
          "usefulness": 8
        },
        "meanScore": 8.3,
        "justification": "Fonts, palette, theme id, 11→9 pages and preserved content match captured tool responses; the spacing explanation and applied styles are also confirmed. Callout Emphasis is plainly unresolved, with concrete remap/add-style next steps. Source labels point to files rather than tool outputs, making the evidence trail misleading.",
        "artifactSha256": "3d769b3122680ad31c5c81c3d8c0acba7b5e516e94c4ce086a746ee1da4e756b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/theme-review-ornith-r2-2026-10-21T2142.json"
      }
    },
    {
      "trialId": "craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-powerpoint-deck",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:45:19.227Z",
      "finishedAt": "2026-10-21T21:53:32.409Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-powerpoint-deck passed 17 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 493182,
        "timing": {
          "startedAt": "2026-10-21T21:45:19.227Z",
          "finishedAt": "2026-10-21T21:53:32.409Z",
          "timeToFirstArtifactMs": 289989,
          "timeToLastArtifactWriteMs": 487525,
          "timeToFirstTokenMs": 256644,
          "firstTurnTtftMs": 10783,
          "timeToFirstToolCallMs": 257118
        },
        "toolCalls": 49,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 97.75,
        "prefillTokensPerSec": 1672.58,
        "decodeTokens": 15573,
        "peakRssMb": 35977,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T21:59:11.024Z",
        "trialId": "craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck",
        "scenarioId": "craftbook-powerpoint-deck",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/workspace/powerpoint-grounding-eval/powerpoint/eval/deck.md",
        "artifactBytes": 1299,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "source-acquisition": 9,
          "outline-parity": 9,
          "grounding": 7,
          "convertibility": 9
        },
        "meanScore": 8.5,
        "justification": "Source was read and cited before the locked outline; all seven H1 titles match in order. Supplied metrics and three actions are preserved; identical seven-slide PPTX copies were converted, previewed and saved. Unsupported clean-signal/no-catalog-noise claims, trust harm and an unrequested extend-or-standardize decision weaken grounding.",
        "artifactSha256": "d3bd4a45b4ec9ab9ec8d480a8f045137c60846e8ad279c42aa9ed7b7f61f1534",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/powerpoint-review-ornith-r2-2026-10-21T2157.json"
      }
    },
    {
      "trialId": "wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "wikipedia-research-brief",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:53:43.775Z",
      "finishedAt": "2026-10-21T21:59:40.892Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "research brief is tool-grounded and passes all content gates (word-band, ordered-sections, menabrea-1842, lovelace-1843, notes-a-g, bernoulli, babbage-design, not-completed, source-citations, no-decoy-citation, claim-calibration, no-false-completion, mcp-research-provenance)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 357117,
        "timing": {
          "startedAt": "2026-10-21T21:53:43.775Z",
          "finishedAt": "2026-10-21T21:59:40.892Z",
          "timeToFirstArtifactMs": 331181,
          "timeToLastArtifactWriteMs": 354781,
          "timeToFirstTokenMs": 251808,
          "firstTurnTtftMs": 8071,
          "timeToFirstToolCallMs": 262667
        },
        "toolCalls": 9,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 105.68,
        "prefillTokensPerSec": 1534.48,
        "decodeTokens": 8540,
        "peakRssMb": 30275,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T22:01:04.336Z",
        "trialId": "wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief",
        "scenarioId": "wikipedia-research-brief",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/workspace/analytical-engine-research-brief/research-brief.md",
        "artifactBytes": 5241,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "synthesis": 4,
          "grounding": 5,
          "citationQuality": 6,
          "calibration": 6
        },
        "meanScore": 5.3,
        "justification": "Correct 1842→1843 chronology, Notes A–G, Bernoulli and Engine design are cited; first programmer remains disputed. Sections mostly repeat card facts instead of synthesizing. Claims that Lovelace corrected/extended Menabrea and articulated a landmark vision lack card support; never executed on real hardware overstates lifetime incompletion.",
        "artifactSha256": "ae8f4fd6ff2dcfda3473aee0d9a0e2682686ef6a6bab0fbc5d2513ecf2d85930",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/wikipedia-review-ornith-r2-2026-10-21T2157.json"
      }
    },
    {
      "trialId": "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-week-plan",
        "repetition": 2
      },
      "startedAt": "2026-10-21T21:59:57.654Z",
      "finishedAt": "2026-10-21T23:23:16.577Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "repair-aborted: khishigjargal/5bea3768 exhausted its bounded automatic recovery allowance at the current checked progress; last error: [llama-cpp] aborting — the gezel emitted 3844 characters of prose this turn without calling any action tool. Stop planning. Your next message must START with a single tool call — or, if the work is genuinely finished and nothing is left to do, be ONE short sentence saying so and nothing else. This s",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 0,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 0% of its measured signals."
          },
          "efficiency": {
            "score": 0,
            "ruleId": "E_BOTH_SEVERE",
            "summary": "The trial exhausted its budget while making at least thirty tool calls."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4998923,
        "timing": {
          "startedAt": "2026-10-21T21:59:57.654Z",
          "finishedAt": "2026-10-21T23:23:16.577Z",
          "timeToFirstArtifactMs": 1108059,
          "timeToLastArtifactWriteMs": 4616433,
          "timeToFirstTokenMs": 279578,
          "firstTurnTtftMs": 26840,
          "timeToFirstToolCallMs": 285845
        },
        "toolCalls": 375,
        "autoAnswers": 26,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 56.54,
        "prefillTokensPerSec": 1048.08,
        "decodeTokens": 159849,
        "peakRssMb": 36572,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T23:32:29.114Z",
        "trialId": "craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan",
        "scenarioId": "craftbook-week-plan",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/artifacts/calendar-desk-week-plan-eval/tasks/eval/report.md",
        "artifactBytes": 5988,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "coverage": 0,
          "prioritization": 2,
          "conflict-handling": 0,
          "realism": 3
        },
        "meanScore": 1.3,
        "justification": "All eight meetings, both prep needs and Tuesday’s budget/Priya conflict are absent; no focus blocks are placed. Follow-up dates are preserved, but FU-11 is assigned Wednesday without a pre-demo time. The report candidly leaves the calendar unknown instead of inventing meetings; calendar request capture is empty.",
        "artifactSha256": "f8c7d1de85f7737baa570d2abf509056e9d3d5e2be2e42f6a78e2cd312dc1815",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/week-plan-review-ornith-r2-2026-10-21T2327.json"
      }
    },
    {
      "trialId": "craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-ab-test-readout",
        "repetition": 2
      },
      "startedAt": "2026-10-21T23:23:35.283Z",
      "finishedAt": "2026-10-21T23:35:43.819Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-ab-test-readout passed 19 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 728536,
        "timing": {
          "startedAt": "2026-10-21T23:23:35.283Z",
          "finishedAt": "2026-10-21T23:35:43.819Z",
          "timeToFirstArtifactMs": 362063,
          "timeToLastArtifactWriteMs": 726481,
          "timeToFirstTokenMs": 270578,
          "firstTurnTtftMs": 23485,
          "timeToFirstToolCallMs": 274965
        },
        "toolCalls": 45,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 63.43,
        "prefillTokensPerSec": 981.32,
        "decodeTokens": 21736,
        "peakRssMb": 37934,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T23:43:01.846Z",
        "trialId": "craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout",
        "scenarioId": "craftbook-ab-test-readout",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/workspace/a-b-test-readout-eval/tasks/eval/readout.md",
        "artifactBytes": 3991,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judgeSpecMismatch",
        "comparisonEligible": false,
        "justification": "Declared safety/mapping/signals/containment axes describe mock-service deployment, while the artifact is a statistical A/B readout. Quality scoring is withheld until the evaluator rubric is repaired.",
        "declaredAxes": [
          {
            "name": "safety",
            "description": "Dry-run posture is explicit: no real credentials, clear stop-before-side-effect boundary."
          },
          {
            "name": "mapping",
            "description": "Fake endpoints/commands are mapped completely to the workflow steps."
          },
          {
            "name": "signals",
            "description": "Success, failure, retry, and abort conditions are all defined."
          },
          {
            "name": "containment",
            "description": "The plan cannot accidentally reach a real service from a mock context."
          }
        ],
        "issueType": "grader-rubric-mismatch"
      }
    },
    {
      "trialId": "craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-annotated-bibliography",
        "repetition": 2
      },
      "startedAt": "2026-10-21T23:36:00.226Z",
      "finishedAt": "2026-10-21T23:44:08.447Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-annotated-bibliography passed 9 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 488221,
        "timing": {
          "startedAt": "2026-10-21T23:36:00.226Z",
          "finishedAt": "2026-10-21T23:44:08.447Z",
          "timeToFirstArtifactMs": 272473,
          "timeToLastArtifactWriteMs": 434906,
          "timeToFirstTokenMs": 255184,
          "firstTurnTtftMs": 10015,
          "timeToFirstToolCallMs": 256827
        },
        "toolCalls": 20,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 74.86,
        "prefillTokensPerSec": 1234.8,
        "decodeTokens": 13117,
        "peakRssMb": 34195,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T23:59:43.587Z",
        "trialId": "craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography",
        "scenarioId": "craftbook-annotated-bibliography",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/workspace/annotated-bibliography-eval/bibliography.md",
        "artifactBytes": 8149,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "accuracy": 9,
          "grounding": 5,
          "structure": 8,
          "actionability": 7
        },
        "meanScore": 7.3,
        "justification": "All six citations and seeded figures are preserved in uniform Summary/Evaluation/Relevance entries. Useful design and publication caveats are explicit, but the prose invents author conclusions, rules out willingness as a barrier, and claims Williams produced a checklist. Garcia is incorrectly called oldest despite Williams 2019; some proposed biases are speculative.",
        "artifactSha256": "f19f48af29e47fd01a5d17d022b96675ae6d8541fb9381003a438c6abc1c06df",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/bibliography-review-ornith-r2-2026-10-21T2357.json"
      }
    },
    {
      "trialId": "craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "productivity",
        "scenario": "craftbook-research-to-document",
        "repetition": 2
      },
      "startedAt": "2026-10-21T23:44:17.416Z",
      "finishedAt": "2026-10-21T23:50:08.436Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-research-to-document passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 351020,
        "timing": {
          "startedAt": "2026-10-21T23:44:17.416Z",
          "finishedAt": "2026-10-21T23:50:08.436Z",
          "timeToFirstArtifactMs": 262408,
          "timeToLastArtifactWriteMs": 295771,
          "timeToFirstTokenMs": 252700,
          "firstTurnTtftMs": 8985,
          "timeToFirstToolCallMs": 254545
        },
        "toolCalls": 46,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 132.55,
        "prefillTokensPerSec": 2011.71,
        "decodeTokens": 6641,
        "peakRssMb": 34992,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T23:59:43.588Z",
        "trialId": "craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769",
        "cellId": "r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document",
        "scenarioId": "craftbook-research-to-document",
        "suiteId": "productivity",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/artifacts/word-document-from-content-eval/tasks/eval/report_files/report.md",
        "artifactBytes": 2658,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "grounding": 5,
          "structure": 9,
          "convertibility": 8
        },
        "meanScore": 7.3,
        "justification": "All seeded metrics and three actions appear in a clear summary, table, findings and numbered plan. Claims that leakage concentrates at status silence and that no other metrics were tracked exceed the brief. Real convert/preview/save produced a valid DOCX with matching hash; convert_document omitted the requested autoTemplates:true flag.",
        "artifactSha256": "89aac3a1b112880b861a546ea7802063f2e37c7e932e9214e2af3c924664158b",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/docx-review-ornith-r2-2026-10-21T2357.json"
      }
    },
    {
      "trialId": "fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "fix-squisq-bugs",
        "repetition": 2
      },
      "startedAt": "2026-10-21T23:50:18.411Z",
      "finishedAt": "2026-10-21T23:55:24.387Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "all 5 signals firing including tsc-clean (great-circle-path, graceful-errors, deep-layer-validation, haversine-untouched, tsc-clean)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 305976,
        "timing": {
          "startedAt": "2026-10-21T23:50:18.411Z",
          "finishedAt": "2026-10-21T23:55:24.387Z",
          "timeToFirstArtifactMs": 291081,
          "timeToLastArtifactWriteMs": 301412,
          "timeToFirstTokenMs": 252465,
          "firstTurnTtftMs": 7717,
          "timeToFirstToolCallMs": 253275
        },
        "toolCalls": 14,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 131.54,
        "prefillTokensPerSec": 2180.74,
        "decodeTokens": 6332,
        "peakRssMb": 28704,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-21T23:58:03.460Z",
        "trialId": "fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs",
        "scenarioId": "fix-squisq-bugs",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "dev-craftbook-routing",
        "repetition": 2
      },
      "startedAt": "2026-10-21T23:55:34.480Z",
      "finishedAt": "2026-10-22T00:01:03.063Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "task batch-exporter/1 was started from the bundled \"bug-fix-tdd\" recipe with no new craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 328583,
        "timing": {
          "startedAt": "2026-10-21T23:55:34.480Z",
          "finishedAt": "2026-10-22T00:01:03.063Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 308339,
          "firstTurnTtftMs": 10997,
          "timeToFirstToolCallMs": 311909
        },
        "toolCalls": 3,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 126.21,
        "prefillTokensPerSec": 2065.22,
        "decodeTokens": 1437,
        "peakRssMb": 28905,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T00:13:02.620Z",
        "trialId": "dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing",
        "scenarioId": "dev-craftbook-routing",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-interface-contract",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "interface-contract",
        "repetition": 2
      },
      "startedAt": "2026-10-22T00:01:12.441Z",
      "finishedAt": "2026-10-22T00:07:43.328Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "tsc emit clean + node integration.mjs passed: 6 event(s) across 3 kind(s) routed and counted — team shape: 1 distinct non-meester session(s) (solo)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 10,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 390887,
        "timing": {
          "startedAt": "2026-10-22T00:01:12.441Z",
          "finishedAt": "2026-10-22T00:07:43.328Z",
          "timeToFirstArtifactMs": 381915,
          "timeToLastArtifactWriteMs": 385980,
          "timeToFirstTokenMs": 323550,
          "firstTurnTtftMs": 8953,
          "timeToFirstToolCallMs": 329047
        },
        "toolCalls": 4,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 122.94,
        "prefillTokensPerSec": 2619.43,
        "decodeTokens": 6902,
        "peakRssMb": 29554,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T00:13:02.621Z",
        "trialId": "interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-interface-contract",
        "scenarioId": "interface-contract",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-code-review",
        "repetition": 2
      },
      "startedAt": "2026-10-22T00:07:59.545Z",
      "finishedAt": "2026-10-22T00:15:46.486Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-code-review passed 4 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 7.5,
            "ruleId": "E_ONE_MILD_INTERMEDIATE",
            "summary": "One efficiency dimension is efficient and the other is mildly intermediate."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 466941,
        "timing": {
          "startedAt": "2026-10-22T00:07:59.545Z",
          "finishedAt": "2026-10-22T00:15:46.486Z",
          "timeToFirstArtifactMs": 414843,
          "timeToLastArtifactWriteMs": 428376,
          "timeToFirstTokenMs": 325958,
          "firstTurnTtftMs": 11942,
          "timeToFirstToolCallMs": 326422
        },
        "toolCalls": 16,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 109.02,
        "prefillTokensPerSec": 2026.06,
        "decodeTokens": 10845,
        "peakRssMb": 32092,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T00:29:05.870Z",
        "trialId": "craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review",
        "scenarioId": "craftbook-code-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/artifacts/code-review-workflow-eval/reviews/rev-eval-1/report.md",
        "artifactBytes": 4112,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 7,
          "usefulness": 9
        },
        "meanScore": 8,
        "justification": "Scope→report→verdict completed; report stayed in the drawer and was re-read. Six grounded findings cover both SQL injections, false success, handle cleanup, logging and pagination; safe patterns are correctly cleared. The required brief/payment source reads occurred only after completion, and scope mislabels the commit snapshot as unsaved work.",
        "artifactSha256": "29367219c855199a17a990954861d378de39ac3ee03f5d711a93bcb274fb91b5",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/code-review-ornith-r2-2026-10-22T0027.json"
      }
    },
    {
      "trialId": "craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-api-contract-review",
        "repetition": 2
      },
      "startedAt": "2026-10-22T00:15:58.716Z",
      "finishedAt": "2026-10-22T00:29:10.683Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-api-contract-review passed 6 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 9,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 5,
            "ruleId": "E_MODERATE",
            "summary": "The non-severe efficiency evidence is moderate, including a 20-29 call trace or two intermediate dimensions."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 791967,
        "timing": {
          "startedAt": "2026-10-22T00:15:58.716Z",
          "finishedAt": "2026-10-22T00:29:10.683Z",
          "timeToFirstArtifactMs": 395268,
          "timeToLastArtifactWriteMs": 741262,
          "timeToFirstTokenMs": 363755,
          "firstTurnTtftMs": 10316,
          "timeToFirstToolCallMs": 365236
        },
        "toolCalls": 22,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 77.68,
        "prefillTokensPerSec": 1401.89,
        "decodeTokens": 26587,
        "peakRssMb": 31935,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T00:30:22.688Z",
        "trialId": "craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review",
        "scenarioId": "craftbook-api-contract-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/artifacts/api-contract-review-workflow-eval/tasks/eval/api-review.md",
        "artifactBytes": 10781,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "workflow fidelity": 9,
          "usefulness": 6
        },
        "meanScore": 7.5,
        "justification": "Scope→endpoint review→report→QA loop completed with drawer outputs and re-reads. Report correctly catches retryCount’s type mismatch and undeclared widget GET, and clears session routing/email omission. It misses the unimplemented history GET and declares compatibility/no existing clients from v1.0.0 alone despite proposing a field rename.",
        "artifactSha256": "1fed052b85f7bb09f54807bfce4e33d9262e958a7dc2158adc21ba4f21cd3b6a",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/api-review-ornith-r2-2026-10-22T0027.json"
      }
    },
    {
      "trialId": "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-deep-security-review",
        "repetition": 2
      },
      "startedAt": "2026-10-22T00:29:22.679Z",
      "finishedAt": "2026-10-22T00:36:38.885Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "craftbook-deep-security-review stale no-write repair loop: security-review/REPORT.md had unchanged content (3603 bytes) for 71s — a re-emission of identical bytes counts as unchanged here after repair feedback and newer chat activity; still failing 8/11: security-review/REPORT.md is missing required content /Systemic Themes|root cause|blast radius/ — nothing in its 3603 bytes matches. Add that content.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 7.3,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 72.7% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 436206,
        "timing": {
          "startedAt": "2026-10-22T00:29:22.679Z",
          "finishedAt": "2026-10-22T00:36:38.885Z",
          "timeToFirstArtifactMs": 345518,
          "timeToLastArtifactWriteMs": 367952,
          "timeToFirstTokenMs": 326785,
          "firstTurnTtftMs": 9585,
          "timeToFirstToolCallMs": 327534
        },
        "toolCalls": 48,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 124.33,
        "prefillTokensPerSec": 1428.42,
        "decodeTokens": 6188,
        "peakRssMb": 34494,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T00:44:14.199Z",
        "trialId": "craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review",
        "scenarioId": "craftbook-deep-security-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/workspace/security-maintenance-eval/security-review/REPORT.md",
        "artifactBytes": 3625,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "verification": 8,
          "systemic analysis": 2,
          "remediation": 8
        },
        "meanScore": 6,
        "justification": "Report and JSON trace all three admin defects, clear both safe files, and give matching parameterization, execFile/spawn and requireAdmin fixes. The oracle confirms 3/3 with zero false positives. Findings lack root-cause grouping, blast-radius analysis and JSON theme fields; unauthenticated reachability and SQL write impact exceed the shown context.",
        "artifactSha256": "cd643abda66a0bce3903f8983753f274c250937f5dc6480e44642862c337cf69",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/security-review-ornith-r2-2026-10-22T0042.json"
      }
    },
    {
      "trialId": "large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-large-pr-review",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "large-pr-review",
        "repetition": 2
      },
      "startedAt": "2026-10-22T00:36:53.350Z",
      "finishedAt": "2026-10-22T00:57:22.489Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "complete large-PR review (coverage-120-of-120, coverage-summary, late-defect-citation, request-changes, no-false-missing-api, source-unchanged)",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1229139,
        "timing": {
          "startedAt": "2026-10-22T00:36:53.350Z",
          "finishedAt": "2026-10-22T00:57:22.489Z",
          "timeToFirstArtifactMs": 646358,
          "timeToLastArtifactWriteMs": 1222225,
          "timeToFirstTokenMs": 271371,
          "firstTurnTtftMs": 15190,
          "timeToFirstToolCallMs": 272236
        },
        "toolCalls": 35,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 36.7,
        "prefillTokensPerSec": 608.21,
        "decodeTokens": 27396,
        "peakRssMb": 31197,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T00:58:03.356Z",
        "trialId": "large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-large-pr-review",
        "scenarioId": "large-pr-review",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-codemod-sweep",
        "repetition": 2
      },
      "startedAt": "2026-10-22T00:57:34.971Z",
      "finishedAt": "2026-10-22T01:05:06.687Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-codemod-sweep passed 32 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 451716,
        "timing": {
          "startedAt": "2026-10-22T00:57:34.971Z",
          "finishedAt": "2026-10-22T01:05:06.687Z",
          "timeToFirstArtifactMs": 307524,
          "timeToLastArtifactWriteMs": 448846,
          "timeToFirstTokenMs": 293610,
          "firstTurnTtftMs": 9257,
          "timeToFirstToolCallMs": 295052
        },
        "toolCalls": 68,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 119.37,
        "prefillTokensPerSec": 2022.34,
        "decodeTokens": 12748,
        "peakRssMb": 33149,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T01:14:28.512Z",
        "trialId": "craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep",
        "scenarioId": "craftbook-codemod-sweep",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/artifacts/profile-cards/tasks/1/verification.md",
        "artifactBytes": 2494,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Enumeration completeness": 9,
          "Invariant discipline": 9,
          "Reconciled verification": 9
        },
        "meanScore": 9,
        "justification": "Real pre-edit ripgrep found 11 matches: seven rename sites and four documented exclusions. Definition, barrel export, imports/calls and log string changed mechanically; eight captured files match only the intended substitutions. A real residual scan leaves exactly four excluded mentions; baseline and three post-edit suite runs each pass 3/3.",
        "artifactSha256": "8b22734791fb6d4c7077ea73993e21340ecfee4105c53dfd23104c061c7ca184",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/codemod-review-ornith-r2-2026-10-22T0112.json"
      }
    },
    {
      "trialId": "craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-refactor-module",
        "repetition": 2
      },
      "startedAt": "2026-10-22T01:05:23.877Z",
      "finishedAt": "2026-10-22T01:22:18.367Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-refactor-module passed 30 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 1014490,
        "timing": {
          "startedAt": "2026-10-22T01:05:23.877Z",
          "finishedAt": "2026-10-22T01:22:18.367Z",
          "timeToFirstArtifactMs": 262738,
          "timeToLastArtifactWriteMs": 987256,
          "timeToFirstTokenMs": 255813,
          "firstTurnTtftMs": 8560,
          "timeToFirstToolCallMs": 258096
        },
        "toolCalls": 88,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 74.66,
        "prefillTokensPerSec": 1221.86,
        "decodeTokens": 42968,
        "peakRssMb": 35377,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T01:30:18.774Z",
        "trialId": "craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module",
        "scenarioId": "craftbook-refactor-module",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/artifacts/crewlog/tasks/1/change-notes.md",
        "artifactBytes": 2178,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Guardrail-first discipline": 9,
          "Reversible-stage execution": 5,
          "Genuine deduplication": 9
        },
        "meanScore": 7.7,
        "justification": "Real 14/14 baseline, characterization of quirks and measured duplicate counts precede source changes. Final code has one shared normalizer and preserves public APIs/original tests. But both callers were edited before the next suite run, contradicting the claimed per-stage green receipts and “no deviations”; review repeats that unsupported claim.",
        "artifactSha256": "389b8b4f77a0d2b479ce83d5447412f0a5bb9e8daa07865a43174808a78fb04e",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/refactor-review-ornith-r2-2026-10-22T0127.json"
      }
    },
    {
      "trialId": "craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "developer",
        "scenario": "craftbook-bug-fix-tdd",
        "repetition": 2
      },
      "startedAt": "2026-10-22T01:22:39.863Z",
      "finishedAt": "2026-10-22T01:35:01.342Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-bug-fix-tdd passed 28 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 741479,
        "timing": {
          "startedAt": "2026-10-22T01:22:39.863Z",
          "finishedAt": "2026-10-22T01:35:01.342Z",
          "timeToFirstArtifactMs": 287334,
          "timeToLastArtifactWriteMs": 737215,
          "timeToFirstTokenMs": 271491,
          "firstTurnTtftMs": 9585,
          "timeToFirstToolCallMs": 272380
        },
        "toolCalls": 55,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 117.95,
        "prefillTokensPerSec": 1875.82,
        "decodeTokens": 18307,
        "peakRssMb": 37409,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T01:44:50.584Z",
        "trialId": "craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx",
        "cellId": "r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd",
        "scenarioId": "craftbook-bug-fix-tdd",
        "suiteId": "developer",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/artifacts/orderly-pricing/tasks/1/fix-notes.md",
        "artifactBytes": 2476,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "Repro-first discipline": 9,
          "Fix minimality": 8,
          "Honest verification": 7
        },
        "meanScore": 8,
        "justification": "A real appended boundary test fails 80≠68 before the one-character fix, then the original three tests plus the guard pass 4/4. Production changes only > to >= and package.json is finally restored exactly. Unnecessary package edits and a workspace artifacts/package.json copy are omitted from the notes; no-other-callers claims lack a repository search.",
        "artifactSha256": "84b4699723f92ccd849875b99aaaf8bf6fc67a1e31a0d866decc2a7b64143466",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/tdd-review-ornith-r2-2026-10-22T0142.json"
      }
    },
    {
      "trialId": "craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-find-vs-create",
        "repetition": 2
      },
      "startedAt": "2026-10-22T01:35:18.297Z",
      "finishedAt": "2026-10-22T01:39:43.726Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "task shop-site-pages/1 was sourced from the bundled \"seo-meta-pack\" craftbook with no new local craftbook authored",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.8,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 5,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 50% of its measured signals."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 265429,
        "timing": {
          "startedAt": "2026-10-22T01:35:18.297Z",
          "finishedAt": "2026-10-22T01:39:43.726Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": 252705,
          "firstTurnTtftMs": 10902,
          "timeToFirstToolCallMs": 254235
        },
        "toolCalls": 2,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 129.35,
        "prefillTokensPerSec": 2523.72,
        "decodeTokens": 756,
        "peakRssMb": 28788,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T01:43:05.432Z",
        "trialId": "craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create",
        "scenarioId": "craftbook-find-vs-create",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/workspace/shop-site-pages/pages/index.html",
        "artifactBytes": 168,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-edit-midtask",
        "repetition": 2
      },
      "startedAt": "2026-10-22T01:39:53.766Z",
      "finishedAt": "2026-10-22T01:46:35.666Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "diagnosed the unwinnable gate, rewrote the failing step prompt, grew the recipe with a verification step, and completed the task",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 7,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 0,
            "ruleId": "B_RED_FLAG_OR_MISSING_ROLE",
            "summary": "At least one red-flag behavior fired or an expected role was missing."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 401900,
        "timing": {
          "startedAt": "2026-10-22T01:39:53.766Z",
          "finishedAt": "2026-10-22T01:46:35.666Z",
          "timeToFirstArtifactMs": 340247,
          "timeToLastArtifactWriteMs": 387965,
          "timeToFirstTokenMs": 253876,
          "firstTurnTtftMs": 13562,
          "timeToFirstToolCallMs": 254781
        },
        "toolCalls": 31,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 100.33,
        "prefillTokensPerSec": 1675.36,
        "decodeTokens": 10033,
        "peakRssMb": 34592,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T01:58:14.785Z",
        "trialId": "craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask",
        "scenarioId": "craftbook-edit-midtask",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-invoice-run",
        "repetition": 2
      },
      "startedAt": "2026-10-22T01:46:47.348Z",
      "finishedAt": "2026-10-22T01:54:02.560Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "craftbook-invoice-run passed 10 deterministic craftbook checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 435212,
        "timing": {
          "startedAt": "2026-10-22T01:46:47.348Z",
          "finishedAt": "2026-10-22T01:54:02.560Z",
          "timeToFirstArtifactMs": 270270,
          "timeToLastArtifactWriteMs": 425758,
          "timeToFirstTokenMs": 253418,
          "firstTurnTtftMs": 9262,
          "timeToFirstToolCallMs": 254813
        },
        "toolCalls": 48,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 140.82,
        "prefillTokensPerSec": 1925.83,
        "decodeTokens": 14334,
        "peakRssMb": 36006,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T02:01:09.435Z",
        "trialId": "craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run",
        "scenarioId": "craftbook-invoice-run",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/workspace/fieldnote-studio-office/report.md",
        "artifactBytes": 701,
        "artifactLimitBytes": 20480,
        "rubric": "scenario",
        "advisory": true,
        "status": "judged",
        "scoreAxes": {
          "numeric integrity": 9,
          "numbering continuity": 9,
          "professionalism": 6,
          "completeness": 4,
          "print-polish": 5
        },
        "meanScore": 6.6,
        "justification": "Amounts 1840+2600+975=5415 and sequence 042–044 are correct; due dates match Oct 21 net terms. Books remain at 041 with all three uninvoiced, despite “went out.” Two invoices omit issue dates; all omit currency. A4 prints fit one page, but 042/043 have large gray blocks and 044 totals misalign with the amount column.",
        "artifactSha256": "23930c5f346250207b58d7dbdd0a31081bce054fa88ca72d59f51a27b435fe52",
        "artifactClipped": false,
        "evidenceMode": "Final artifact source read against exact frozen axes; selected seeded inputs, saved source and companion deliverables inspected. No external judge call.",
        "reviewEvidencePath": "evals/runs/scorecard-seven-models-2026-10-15-proposal/invoice-review-ornith-r2-2026-10-22T0157.json"
      }
    },
    {
      "trialId": "craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-linear",
        "repetition": 2
      },
      "startedAt": "2026-10-22T01:54:13.868Z",
      "finishedAt": "2026-10-22T02:42:51.413Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "authored a gated 3-step craftbook, invoked it, ran the task to completion, and all three outputs pass their checks",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 2917545,
        "timing": {
          "startedAt": "2026-10-22T01:54:13.868Z",
          "finishedAt": "2026-10-22T02:42:51.413Z",
          "timeToFirstArtifactMs": 1329964,
          "timeToLastArtifactWriteMs": 2897818,
          "timeToFirstTokenMs": 253598,
          "firstTurnTtftMs": 13690,
          "timeToFirstToolCallMs": 255079
        },
        "toolCalls": 96,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 67.73,
        "prefillTokensPerSec": 1054.06,
        "decodeTokens": 65788,
        "peakRssMb": 37858,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T02:43:15.251Z",
        "trialId": "craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear",
        "scenarioId": "craftbook-author-linear",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-route-multi",
        "repetition": 2
      },
      "startedAt": "2026-10-22T02:43:02.869Z",
      "finishedAt": "2026-10-22T02:50:08.855Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "invoked the bundled press-release recipe, ran the task to completion, and the release carries a dateline, the attributed approved quote, the boilerplate, a contact, and the approved facts rather than the stale draft",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 425986,
        "timing": {
          "startedAt": "2026-10-22T02:43:02.869Z",
          "finishedAt": "2026-10-22T02:50:08.855Z",
          "timeToFirstArtifactMs": 289667,
          "timeToLastArtifactWriteMs": 360215,
          "timeToFirstTokenMs": 250552,
          "firstTurnTtftMs": 10851,
          "timeToFirstToolCallMs": 253023
        },
        "toolCalls": 50,
        "autoAnswers": 1,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 123.9,
        "prefillTokensPerSec": 2117.17,
        "decodeTokens": 14405,
        "peakRssMb": 37011,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T02:58:12.809Z",
        "trialId": "craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi",
        "scenarioId": "craftbook-route-multi",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-export-generalize",
        "repetition": 2
      },
      "startedAt": "2026-10-22T02:50:20.192Z",
      "finishedAt": "2026-10-22T02:59:22.339Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": false,
        "reason": "chat stalled for 303s — daemon reachable but issuing no model turns (soft digest stuck at 7a15590c247174ba)",
        "failureMode": "chat-stalled",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 3.5,
        "band": "framework-gap",
        "axes": {
          "completion": {
            "score": 0,
            "ruleId": "C_WATCHDOG_LOW_OR_NONE",
            "summary": "A watchdog ended the trial without evidence that at least half of the gate was reached."
          },
          "quality": {
            "score": 0,
            "ruleId": "Q_NO_PRODUCED_OUTPUT",
            "summary": "No authored artifact event or scored sniff output was observed."
          },
          "efficiency": {
            "score": 10,
            "ruleId": "E_LOW_BUDGET_LOW_CALLS",
            "summary": "The trial used at most half its budget and no more than ten tool calls."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 542147,
        "timing": {
          "startedAt": "2026-10-22T02:50:20.192Z",
          "finishedAt": "2026-10-22T02:59:22.339Z",
          "timeToFirstArtifactMs": null,
          "timeToLastArtifactWriteMs": null,
          "timeToFirstTokenMs": null,
          "firstTurnTtftMs": null,
          "timeToFirstToolCallMs": null
        },
        "toolCalls": 0,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 0,
          "clamp": {
            "requestedTokens": 65536,
            "grantedTokens": 1
          }
        },
        "decodeTokensPerSec": null,
        "prefillTokensPerSec": null,
        "decodeTokens": null,
        "peakRssMb": 916,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T03:00:24.433Z",
        "trialId": "craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize",
        "scenarioId": "craftbook-export-generalize",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-params",
        "repetition": 2
      },
      "startedAt": "2026-10-22T02:59:35.071Z",
      "finishedAt": "2026-10-22T04:19:44.803Z",
      "includedInScorecard": true,
      "scorecardRunId": "2026-10-21-m5max-ornith-observed-subset",
      "outcome": {
        "success": true,
        "reason": "authored one parameterized craftbook, ran it twice with different region inputs, and both region summaries exist with distinct contents",
        "failureMode": null,
        "failureClass": "pass",
        "includedInFixedModelAggregate": true,
        "failureClassRule": null
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 8.5,
        "band": "ship-ready",
        "axes": {
          "completion": {
            "score": 10,
            "ruleId": "C_SUCCESS",
            "summary": "The scenario reached a successful terminal verdict."
          },
          "quality": {
            "score": 10,
            "ruleId": "Q_GATE_COMPLETE",
            "summary": "The scenario gate is complete."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": null,
      "performance": {
        "durationMs": 4809732,
        "timing": {
          "startedAt": "2026-10-22T02:59:35.071Z",
          "finishedAt": "2026-10-22T04:19:44.803Z",
          "timeToFirstArtifactMs": 1663823,
          "timeToLastArtifactWriteMs": 4805140,
          "timeToFirstTokenMs": 301712,
          "firstTurnTtftMs": 47758,
          "timeToFirstToolCallMs": 306355
        },
        "toolCalls": 123,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 45.79,
        "prefillTokensPerSec": 843.79,
        "decodeTokens": 90212,
        "peakRssMb": 37618,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T04:28:19.708Z",
        "trialId": "craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params",
        "scenarioId": "craftbook-author-params",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    },
    {
      "trialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st",
      "cell": {
        "id": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "provider": "llama-cpp",
        "model": "ornith1.5-35b-a3b-q4",
        "targetId": "ornith1.5-35b-a3b-q4",
        "suite": "complex-work",
        "scenario": "craftbook-author-gate-script",
        "repetition": 2
      },
      "startedAt": "2026-10-22T04:20:00.702Z",
      "finishedAt": "2026-10-22T05:42:18.657Z",
      "includedInScorecard": false,
      "scorecardRunId": null,
      "outcome": {
        "success": false,
        "reason": "retry loop (stall-path): sniff \"craftbook-author-gate-script:4:targetnone:fr1q7wvh:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 5 tool calls (0 re-writes). Artifact produced but never reached success.",
        "failureMode": "model-stuck",
        "failureClass": "model",
        "includedInFixedModelAggregate": true,
        "failureClassRule": "model-default"
      },
      "fixedScore": {
        "rubricVersion": "eval-run-fixed-v1",
        "composite": 5.2,
        "band": "needs-tuning",
        "axes": {
          "completion": {
            "score": 3,
            "ruleId": "C_WATCHDOG_PARTIAL",
            "summary": "A watchdog ended the trial after at least half of a measured gate, or after observable progress with an unavailable denominator."
          },
          "quality": {
            "score": 8,
            "ruleId": "Q_GATE_RATIO",
            "summary": "The latest scenario gate reached 80% of its measured signals."
          },
          "efficiency": {
            "score": 2.5,
            "ruleId": "E_ONE_SEVERE",
            "summary": "Exactly one efficiency dimension reached the severe threshold."
          },
          "behavior": {
            "score": 10,
            "ruleId": "B_CLEAN_LOW_INTERVENTION",
            "summary": "No red flags or missing roles were observed, with at most one intervention."
          }
        }
      },
      "comparisonIssue": "Frozen comparison changed: fixtures: \"auto-answer.ts\", \"batch.ts\", \"bin/ab-work-routing.ts\", \"bin/image-bench.ts\", \"craftbooks/adapter.ts\", \"craftbooks/authoring/author-linear.ts\", \"craftbooks/authoring/edit-midtask.ts\", \"craftbooks/authoring/export-generalize.ts\", and 40 more; runtime: \"@bendyline/gezel-mcp/package.json\", \"@bendyline/gezel-service/dependency/@bendyline/docblocks-cli\", \"@bendyline/gezel/dist/browser.js\", \"@bendyline/gezel/dist/chunk-3Z6PBVZM.js\", \"@bendyline/gezel/dist/chunk-5BXY37IX.js\", \"@bendyline/gezel/dist/chunk-6ASB7JVJ.js\", \"@bendyline/gezel/dist/chunk-6RY2SMVG.js\", \"@bendyline/gezel/dist/chunk-6YKM4W6M.js\", and 28 more; source: \"evals/src/auto-answer.ts\", \"evals/src/batch.ts\", \"evals/src/bin/ab-work-routing.ts\", \"evals/src/bin/image-bench.ts\", \"evals/src/craftbooks/adapter.ts\", \"evals/src/craftbooks/authoring/author-linear.ts\", \"evals/src/craftbooks/authoring/edit-midtask.ts\", \"evals/src/craftbooks/authoring/export-generalize.ts\", and 40 more (see identity-drift-after-trial.json)",
      "performance": {
        "durationMs": 4937955,
        "timing": {
          "startedAt": "2026-10-22T04:20:00.702Z",
          "finishedAt": "2026-10-22T05:42:18.657Z",
          "timeToFirstArtifactMs": 2037692,
          "timeToLastArtifactWriteMs": 3619387,
          "timeToFirstTokenMs": 253585,
          "firstTurnTtftMs": 13662,
          "timeToFirstToolCallMs": 255093
        },
        "toolCalls": 82,
        "autoAnswers": 0,
        "engineContext": {
          "launches": 1,
          "grantedPerSlotTokens": 262144,
          "totalTokens": 262144,
          "slots": 1,
          "kvCacheType": "f16",
          "launchModel": "ornith1.5-35b-a3b-q4"
        },
        "decodeTokensPerSec": 36.87,
        "prefillTokensPerSec": 444.45,
        "decodeTokens": 75881,
        "peakRssMb": 40572,
        "gpuTelemetryAvailable": false
      },
      "source": {
        "result": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/result.json",
        "score": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/score.json",
        "facts": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/facts.json",
        "metrics": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/metrics.json",
        "inlineJudge": "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/codex-inline-judge.json"
      },
      "judge": {
        "judgeProvider": "codex-inline",
        "judgeModel": "current-session",
        "judgedAt": "2026-10-22T05:44:11.191Z",
        "trialId": "craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st",
        "cellId": "r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script",
        "scenarioId": "craftbook-author-gate-script",
        "suiteId": "complex-work",
        "modelId": "ornith1.5-35b-a3b-q4",
        "artifactPath": null,
        "artifactBytes": null,
        "artifactLimitBytes": 20480,
        "rubric": "none",
        "advisory": true,
        "status": "unjudged-no-declared-rubric",
        "comparisonEligible": false,
        "justification": "This task declares no qualitative axes and is not an HTML anchor; its deterministic trial score remains available."
      }
    }
  ],
  "evidence": {
    "availability": "References and SHA-256 hashes only; original artifacts, full transcripts, host paths and credentials are not bundled.",
    "pathConvention": "Repository-relative paths refer to the retained local evidence, not files shipped in this publication package.",
    "sourceFiles": {
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/campaign.json": "54a270ad565e365a92b9e969257d6951a7cc68e2136ab0ebaf902379bd684939",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/campaign.json": "e19b1aec94dcd5e572564549762b56731b38d08a60d4e29a2f1a17b70881ea68",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/campaign.json": "ca3cd56fbbfcc1716cda9350ab85d62474ae431b3486274f3ff9dc4578684468",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/campaign.json": "66488f9f31e8a3f01dbd6f4cef1d6ffd1917b5e5fc54a1ed7dcc46f4279a254f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/campaign.json": "272af8d883374dc96e9ef91a457fe50ef6b7c9ee070bddc143248fd6886d0143",
      "evals/runs/scorecard-seven-models-2026-10-15-proposal/final-report-601-trials.json": "35ac14b8a537f567a26364b3537e788fdc250787ffa49b2947980f72337f0a28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/codex-inline-judge.json": "c2a8e5b3a39b4df804fb3a6e229574d597a6606e3c33edcff3375146a94bb333",
      "evals/runs/scorecard-seven-models-2026-10-15-proposal/judging.json": "220a9a2adc5b2148083c5dcf5fbc9120e68e97a342e609187be717da4ff1d8ac",
      "evals/runs/scorecard-seven-models-2026-10-15-proposal/shared-baseline-identity.json": "865f510e47cfac0de3e01dc340cdc7cd190c67e064ef8c24ff5ba6bb7961ecbb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/identity-drift-after-trial.json": "a5ddd7df9683e9cba29b4271293af9a86b928253aa0e8ae5ff31b74f2b8b63a9",
      "evals/runs/scorecard-seven-models-2026-10-15-proposal/identity-review-2026-10-22T0542.json": "5bb7a8f80239ecf81891b62b47ce696dca8fb89431dd00d8688bc346f153ca5a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/judging-issues.json": "541c2e4ce918336a426d8f9370346230e9713156f7ed39a0a33e992890309568",
      "evals/runs/scorecard-seven-models-2026-10-15-proposal/monitor-state.json": "0e3761a6ebdebd1e0e9775665e3a4dfb253b738f9f5dd59f6c16caffae614ad9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/preflight/preflight-lfm2-5-24b-a2b-q4-2026-10-16T04-34-25-418Z-oc39/preflight-report.json": "965391a69acfdc424caae3d62ffe2a615f61fa9c7d9e3ff22ba46e5968418ada",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/preflight/preflight-lfm2-5-24b-a2b-q4-2026-10-17T05-03-19-676Z-tiim/preflight-report.json": "ac71559b5d4f90af28563524fdad3ea499d78e9440b1d8820cd78afb4dadaa3b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/preflight/preflight-qwen3-6-35b-a3b-q4-2026-10-17T08-23-07-242Z-67fj/preflight-report.json": "b29ae28647555260095d8a1b10eba69fbc116ce9ca643e9f6ec6b8365b4cc75c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/preflight/preflight-qwen3-6-35b-a3b-q4-2026-10-18T08-43-09-123Z-woji/preflight-report.json": "06e3ca70d1bc3989cec195c9048e2cd1524a38a5901751a91455b99146ab8f24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/preflight/preflight-qwen3-8-27b-q4-2026-10-18T11-04-38-925Z-chph/preflight-report.json": "1b7ea229fd18d4bd448adcc42f3493cd7ed5f164b4973d6677fe8b402625ce09",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/preflight/preflight-qwen3-8-27b-q4-2026-10-19T11-33-09-990Z-7cpf/preflight-report.json": "e1fdcb0c1444a0c8d1831818bf7c7aa89b835ff36f34f3ec2884c42207d4ca13",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/preflight/preflight-btl4-35b-q4-2026-10-20T00-48-49-039Z-5xna/preflight-report.json": "88884372ead89ceb0f956c750bd4f662b74f0316f8eb16cd2e886af884919357",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/preflight/preflight-btl4-35b-q4-2026-10-21T00-57-14-888Z-m29i/preflight-report.json": "de6ba3de81981869138b871bd4823c12c1f1600ee3ad9bee6d764f8cb783de22",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/preflight/preflight-ornith1-5-35b-a3b-q4-2026-10-21T09-49-24-537Z-jnst/preflight-report.json": "816eee077dd14d16eb856964e37211d758325da2ef99aac69632bd144f6c0f25",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/result.json": "3421aa60881fea219d64c79677f8f36b6ecbac271d2b73fc871aae3a29977d99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/score.json": "ebb13bc89c8499bbaced0b745139aa176ee9f77eaa317eaebff79f64c9fc5054",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/metrics.json": "3dd1e1e8c742e9850103ff132d65e7684aa44ab089c1132753d192460ddf817d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/facts.json": "7a9c3be769c75da49f8e3171d66cd940a1ca503cfd56e0b0418b3a6d913768ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T04-34-42-564Z-kvt1/codex-inline-judge.json": "4074ec5350399689fedf0d100ba680e7198251b30e7a29cc23e87396e205bb5f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/result.json": "632b904ce87664c145cde376dc1953295126dc6a3d2defffbce51acda1a222c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/score.json": "80bf147b9714fa273874be1039b91493527e86089dff7ddaf04a4d39f8fadf16",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/metrics.json": "c8d49447b8eeda83bb9bedd374575ca9f07658be726bc4bf27aae09b2e8adb03",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/facts.json": "0ef184ef0289146a57f3ea082fc9ccb7464c478780f9343ad97d16efe6ff47c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T04-36-08-035Z-35ki/codex-inline-judge.json": "1400173c8e07d9a96d065f09b4d7103fc407f1b0f2434990032a75ef7da89206",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/result.json": "5791f0b0db50ea6d4980e778df7211f22f486e5f3ac3fc3919e622d22e2eeda7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/score.json": "ffa924a813136ebc5202d059596344f7a3407d2f0cfacff4e35f9dec5ced8690",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/metrics.json": "aea793312c1bb7863bb7833419b9cda3a1cf8e9b39c9fcee3956ae2a811657fb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/facts.json": "239fffc637f08cfb12cb5f5a248107ef78ea1d4500ce987af31c86a94be5f262",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T04-38-18-976Z-wghr/codex-inline-judge.json": "ed20c66736786fe9e3b9ec9d602bb9a44409a8dc0e21a9eff3ae303a66affbc0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/result.json": "cc82ac61a3ff6459e80b7eb9c07318afb08c34f476790f02dc99880a2bb1735c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/score.json": "40eda280a07ee7730b4d4d78cef27bea822d2752970d236f2b9e2173f3ae9836",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/metrics.json": "b1e9140394c33517bb5a9c74e06d6def3ff2978bf475f345c87240f1ca38a995",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/facts.json": "48dde0afefad04295f97c75317fd46129696864a48be8f3e0e8219ffe454a3c4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T04-48-21-033Z-vpws/codex-inline-judge.json": "e28d781a521f553a4e373b38d4310afcca1d866dda921e5805bb01d9510a6018",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/result.json": "f071430804b860c65f602aafe5ebd7885e8781f0ede053ef32f96b7712a1a7fe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/score.json": "ba573de6cfbcb1216e1239345c9e80544a4af7d5cacff5ab3ff560da38d20057",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/metrics.json": "efd4d321acdc0472312e0e8bd39e4fe39493bd737769412099a59a0c8b1758a0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/facts.json": "0391f63b21e8e8bf2848ef9cd72af110779dc55885b1986f610d9a2fd55dbb05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T04-50-55-139Z-zbx8/codex-inline-judge.json": "01f8e72eab004aa2e80da94a46b8143bb4735ccf8135cb1beedbc83e4f6884c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/result.json": "da39c312d3e9a2493e1abf26aa2f48b71fa0bb4d7abd6acaf2363e5fa6b947de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/score.json": "e0b472b43648e075618874d2a02a4eda75a36b3124af28553e494dd831b00828",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/metrics.json": "5d0e314f4e7c5947044ac4ed922e83f39213a6b55686e6908a62ce2cd913835d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/facts.json": "de68bb53741b1dab1f1930540bd57872dec44a31ee38ff512288bd12df03fb67",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T05-04-54-299Z-fejd/codex-inline-judge.json": "9bd5354a79e438e7c5bc8989a65cb0191ea962f096896fee192f708a5fecd682",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/result.json": "0015fc588f14e6ef8da80fd5da644fb35109e4fcdfabeeff542a6f920c39d7aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/score.json": "c8a0cd5a264e484c05f0251fe5ddf69adb8271758185df238909c5608363718d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/metrics.json": "cf3beafd036b9d1d5a7bb3574e9195e5c93f5a18390515424af4a30d895d399f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/facts.json": "69b750f33eb5e55055155c6126e9d7798474dd692e0869336d8ee60d9f0f495f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T05-06-06-733Z-5e3b/codex-inline-judge.json": "4fec0e30f4faa1c04146d115295af37d8be84a1f7aad01c8dcb94ea1375c8bb5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/result.json": "88fcf193b86b016d44cd5844824de0e2f43f053ecee88f43b7beb7c7d4aba3b5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/score.json": "15f0f2eba2234a17646a094b4b485a5d6174beec164f00617ead563810c690c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/metrics.json": "4b103f8c648b567d22fa3edef670bd8c467bae50584aee85617995239c28fb96",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/facts.json": "b3ea61ddd1912cdae41430c3e57939ae27dd3363080f9a3efa8cd6c4aaeaf339",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T05-12-07-451Z-3lib/codex-inline-judge.json": "2cd53f973694be0455b411a5df82eeebfb739f857560af15b3aa50599d7f8adc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/result.json": "990d1d3e9dab974a6fec77177f04196881c000374c45622f710a17506923e52e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/score.json": "f73568c022aa2eb0f09e1016f2553220b70fdde10e910266c0e13e5d9a1206bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/metrics.json": "7c8fb337dea114a0f9a69228e495a24511948a58f3290d4d0be3fe5905f85a56",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/facts.json": "f833e289d1006e21c8caa8cec351358ed7b3b3b69bfaddb37c8c9e99fe2d4405",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T05-14-50-099Z-ak6z/codex-inline-judge.json": "91df75b4ca1fd61626f85cfaea553fee857adc22bdc6cee21569967e835729f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/result.json": "7158083d61e47e8a30ba2fe00ceb5f6c93bfe5c13111c3870101e6bb6298de7c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/score.json": "7f6f6f5628fcf77200cd7d009ae788e6f24c12c786e4dc5beeb55b43beb1c25f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/metrics.json": "9978e83a60234a783791581c77d4a834308d9fb9ad25390bc88b84ccabaa4594",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/facts.json": "5ac90bde5ffdc47435846abb501abd61ea8164d2ddba3e56d3c0934c7415b1c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-15-36-306Z-yvjb/codex-inline-judge.json": "dd934cf47e16f7fe7766b22f62acd5344f278c5e0bebdf206608b4caed0af765",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/result.json": "679c940eaf221b76d0b44f12464d94a7e956b4122af26f7ce96b0ad96fec6d72",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/score.json": "fdbee7be7749216ab85ae707d4805e9418320cd3a5298b626f3d6d6837add5e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/metrics.json": "e7e7675745c5d91640273642859e9706d8a44b10292d09235f4c02919d4a6bc9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/facts.json": "2ab75277b0c1644404c3ad8717ba3f421c6382037765caa49f1dd199e019ff34",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T05-16-48-726Z-36nb/codex-inline-judge.json": "902fe2987de9570526db46be5a26cbc041a72212244c01401bfda02153bf30c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/result.json": "3f0d99581a44203a023e558f73d76639d47466396e5742945a094bcb663cc72c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/score.json": "037fa45de319f52e0439a08c2e2b7ad584b0e451724ac834f1330c8b803ec8bf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/metrics.json": "36f5f8b52a0fb01a543c91b779de87711e00fc4cfa2b21ad35b7ed029578d14b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/facts.json": "6fd8a056f13b3119461fc668bf398762f375a5740ee22b82251ebf27fe62dfea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T05-23-09-503Z-03q7/codex-inline-judge.json": "7b7d3dd884c1c6b9b5c5c0d42c1d1498dbcb90e0be0b44b19d08f9044ecff257",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/result.json": "e96675f7e21265905b16cd6fbb5644b7bb33f01fa9910a07e381be2255157d7d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/score.json": "b003faac33bcfce54865591efd263dcef876b49e69691a61db856ab88a4dd6f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/metrics.json": "f48ba6b0d8752f2196b11cf4bbd006ae9d356d47b27c1cd00da532e66dc7f246",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/facts.json": "d17e0312551b8ea2ea3b9356b4e74389f7d230c09ebcc83d561ec07698c6d936",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T05-24-35-039Z-by3i/codex-inline-judge.json": "fe7b59c96e430b8b9ca5b4047d6b50fa8357076c25870bbe84e6f0fc07d2859b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/result.json": "722b8fc85e30458acbfd6ddaa61e5787611f0ba0e9dc5cbfcb525447d8a92ded",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/score.json": "edc155017f180d25a45b650b52ea0d1a4c2e07dd898f78eaf3688ba802c58cc4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/metrics.json": "e717b227b344ba8f2c8959be6866c2ff1b37fb8e55ce92db0a5be90da56ed109",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/facts.json": "73261956be70120e7fd97b628726be7f0e75b74f8ef8cdb6658b0a2376686628",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T05-27-48-225Z-166h/codex-inline-judge.json": "b45b33df39a326bf36fdcccd1091b09e8ca33eea1e180ad1f861b1797eef968d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/result.json": "6095f6bf641023624a8c75f488b7b59a634c47e0601b8c328b0e6eedb55a62b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/score.json": "c6427e0749684cf2acf2c5cb5bb51fc05e4df5bbf0aa49d0135086c873db3bb0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/metrics.json": "ef7674e6fdfc02d23c9a3dab71b3a2d24ef67067c7c442838fc177459c153991",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/facts.json": "b29063be7d8ba2e5392dc9e1f75e56244199aae7f6a2deef7153254394a1c2cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T05-29-35-433Z-zxut/codex-inline-judge.json": "f359a23fc7e59d80ce395fc76c0537c5cfde7016fea52b50353fb159a3993c30",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/result.json": "c69d717403145772fe3f8c316a71ea45d772582cf1c6a10ff1c224cd93452fdf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/score.json": "edf5c694c601b44633ee46c837d4d15719c03e0064317e5c34fc3a6cc4a49f5f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/metrics.json": "2ae42d4babc870a8aedf1d00879ee7d950e12678efd0f8f7a34946b3ec99c2f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/facts.json": "0992065d16f59c225eb728a027ff0087ba552bc54592d99f5f3d95efc05542e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T05-33-06-868Z-f39n/codex-inline-judge.json": "c09b1f12aeda3ff9dde4d8efbe18f36d448f4c102908d1ceac4acbe1b66480d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/result.json": "c4f82453469e5688476800792b0d90028df394d66b8e53e6c7cd107f7aba2bfe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/score.json": "2bac5dcee6feb39e5e63ffc288bf4e7064751fc791b58a12d03ed834b0ac68b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/metrics.json": "b5e53ea00b41c707d0e9c329d208adf56bcbf0b2403e69320233bd2566cfeeca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/facts.json": "906e2373ca328b976c81fb9c073943ac7cc7dc53076bcd95b01bb4e2c6ea74d3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T06-02-32-538Z-ga1e/codex-inline-judge.json": "90fad5bf28107c189374cde6f74286b1b41ffeb675fa26b8f169075ce9fe1c60",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/result.json": "90e41cf3bbb3e051578e327f18b3e0c72394fc5732122057be6131a74e298844",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/score.json": "6f577c28841f211504a0130b61f850c2f7412aed3fd8f204be002e5ea6f03122",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/metrics.json": "e20315359ab96a9e2746edbff3b6a7e8eea5a57ade1f68c0d8e04b294fbafaa6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/facts.json": "8c54a73ab2d15a77a0454cdf7e3b6ce296444ae42f0452497df080703acd191f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T06-08-21-745Z-tc1g/codex-inline-judge.json": "112420be977ef6777140db2a6836c5ee6eb58f268dc76f4b29ca63bcabe29af9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/result.json": "a13c89c3bf652382ecc35e4c57af569079a5c9dc4c20679245bb9b345022a0d3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/score.json": "52ad3ce9490b3dbc99a8c3d4e4f1ee665ef9789eda5b6f449c2d7b7c499eda0f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/metrics.json": "a471c8cfda02d52635b6ccc78a7f5335b8cd3eba159fcacd19963b6de22351a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/facts.json": "37fc3000b8296673fb1a145879ab534e2e5b00f82d4fdd7fde655a9ef80f2ca9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T06-10-16-466Z-lskh/codex-inline-judge.json": "9b8ff5964449d4f21b1f056095993185248969ccbdada70862252f67d1aee776",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/result.json": "4e991cef85f980f7d661422fdd8b44af10e640103c6a002dcaf486e5dc602572",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/score.json": "57cce1438db23b3337ccb142b251f79a28051be69ed1e26f92c193c880b08c67",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/metrics.json": "6842c12a33bf7bd8cbfc2a3702a84eb3c39ad286eceb02873bf8481251ef57b5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/facts.json": "422adadd2fb639eae0aade1b127ea4a1e0d0a3ffb9ab62c52a97bb37a80c2721",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T06-19-54-490Z-7y5d/codex-inline-judge.json": "fe49dafd7369e734437d3890a4916c3737d337b5d737d486fdc66f288cb6e228",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/result.json": "fcddb85a09be9b04ea1414efe83959c7ac07376077bb35f0fe86076f0fa1d2db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/score.json": "0b6c8da8c6deb46ec9dc8893c38a7656e0da409ed5c11e46abc956b1994efafc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/metrics.json": "092b97d8faff613316431c30ab57ccc19479236cbf3f497c595e7bf6ff33fc0b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/facts.json": "914c0b716311eb4152adec583909171090aa83a5398f91f29863c2cc793501fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T06-26-29-060Z-9qgg/codex-inline-judge.json": "65c736ce50c173fd61f46cb657be285754c9951bc6c24e6bbe41a05b48cbd168",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/result.json": "822c15c1163c581c11e1dc7e951f15ccc1f68feff1813115d0748f501d8be473",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/score.json": "b499c5ba362973819721924c941e176666b752210d3953cacdeaa4f0f672e4e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/metrics.json": "9bcc054ae9d85e4efaa6ade89edbe767f2137cd01b71ca3f2f9595a41ea94959",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/facts.json": "cef61ba9f12f7c48150faf7588a194dae7ab8e4ceae6c9ca74fd9ab279c3556d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T07-56-50-607Z-j1to/codex-inline-judge.json": "1d6ec322365f8e4e13a638b6778a0a05b9700d8f78595064440412f62300291d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/result.json": "9a9e069ce4be848fc610ca340f970d793090c1f1d5020ac13803925d2e318c63",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/score.json": "945fefc45a18e922fa60bb06c4fc5fd24ceeb9be2bf4306fbbda3bf86d265b83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/metrics.json": "52ce2c5c69ce989ab4af2364d9fd56cd9ecc3f96552562511ec23e63bd772644",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/facts.json": "f31c6096de61065a952004c709c0a6eaa54fe6e28b459193bf5a09e17f9f7827",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T08-08-40-565Z-cn6r/codex-inline-judge.json": "60c42a9c325ff26c6f27ef352651f35fbb9040dd2618513e5be91435b45cf3b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/result.json": "8868ecdd646b6ed44c6d51d1cf9839d5c521c22b09d3edacb44852a2b2626cda",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/score.json": "d43f3993bb40d8055ffd964109f1f08144c0f8cbacde73a837d4349c840c9952",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/metrics.json": "1b3e39f97cfd4bafc8e3d6d210ff0ed21c2910b6515fa63fe57a36fa4eddea24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/facts.json": "679dae30bdc04f32d2e5b769555c0981b9316e9070b7952c06dba989241bbd55",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T08-12-50-708Z-rzdh/codex-inline-judge.json": "2587c78eac1794e49cc5b61c2f3beeeb1eea15071d41602f0d5baab8841129c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/result.json": "786f8aff0a1d53edcd315b5e3ff33ed494c0b0985761ed3e66c9eddb4cedb6cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/score.json": "96a16acdf0370c59b0ed8b913bc104b3a3c5bc114fec87bcaaae4baee32fd778",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/metrics.json": "5e3ceff6f4dfb77096f85ed256ceab81f341376170298a65c69d045a5b402f7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/facts.json": "6d970cdaf08903ce21790c8e5ab52a22a67c5e556daa024ad3bb91c066f31fb9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T08-18-10-225Z-2lzy/codex-inline-judge.json": "4dd27ce95047804464ab4262799d9c4251cefa6a97c9ed2e363374c300091cce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/result.json": "8b4ad086c254987c1f3340333041d64af616b86413eb7f1fa65ecfb899cbc8e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/score.json": "f4813f1acc5500003f72e362e31a3304e27ac9c5be65bc1c5a11fa7f4a24b1ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/metrics.json": "aa61337bb9fa08ea876fc4a3cf70185a051c5521d29a4986a7b238b73abb73ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/facts.json": "9b03555f5a51df7cd051b9baf762f7fcb8f47e14c657e6eeaf2ed93ee684b758",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T08-21-38-262Z-iggl/codex-inline-judge.json": "ad427834229e4a86b05e606285c8df86f9856eaf35986712849f5304a7ca663d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/result.json": "76bc7a8ad4f3b307a2ff2e61f6f6af025f447a681000e0909aed2fbca7c10c09",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/score.json": "ad8ff2997f7e413d104ae1649c88fa8af41ec398329c18ece76e4914d9d08620",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/metrics.json": "018071390d23a0afd0e3020feae072671668f9db98faaed2b45cc57a8936302c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/facts.json": "0db486aef9957ce887a503560f7ff18dc57a7cb2ed2ebc0c143ed30c86548587",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T08-22-25-596Z-7ugn/codex-inline-judge.json": "7886251be398f724cb89fac1bab2aa993cd333865082d37d829c34424f540c72",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/result.json": "68e6793a2ce4a6239cdf3db1d18c8fff0d4fbf8515d2936ec477d18eb42ea3a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/score.json": "00177929f2493056a2f9fc89ef0c545b937b09d87e532da944c242d491e6db3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/metrics.json": "e8aae9302d007655ef14f045d0c08affcdeae20403903bfe76e00be7ef678d34",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/facts.json": "13c1ee596a8370b25fe8096446ec37da30012f8115dd5b7123ed7df69363d1c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T08-27-46-214Z-ftch/codex-inline-judge.json": "e15ea8d40fae8c5ad48f7ea1b65ae024461f407996968a197a6d8ae7e6526d27",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/result.json": "2744104537921144a7f22bf94058d63ba40030688cdee8d32573e8c9e4dee945",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/score.json": "0f1adebdceafbb99d78e847364e0ad22ec761116aff1f4adae79746424abca1f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/metrics.json": "3a8e5ea9930398626f31f8ca83311c1dd9f5cfe9a73f092d9fa14847e12f113c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/facts.json": "6aa7069aca1edd2774f821daeae882e28e6b52177699d4a50bb6b633db7a95bf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T08-48-29-350Z-hxbf/codex-inline-judge.json": "f117b2651842b33b0f54f7bd9192feca6dc28c9512b2b2e2ece7aa0cbc697ee9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/result.json": "54df25cb794a63c949fb3e2c3ea32254e50845f4328591bc644515f8b0fa54b0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/score.json": "73e7cbba817434e245bf11672e5a5f8a6dd7cd11e19d69b6a9d1932eb2aabe45",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/metrics.json": "c0ee7822547498d17b7c700917d70abcde301ce136478f901ecdb8019009ed72",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/facts.json": "35199abdbba23b70db0d14ea7109fff4a20ebe89b696cb360ea0971aa4cd2e03",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T09-04-52-359Z-smhv/codex-inline-judge.json": "176ef630d007dc9a7b6e8d55175c1033a79e751a0c118530475125d3c10de91e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/result.json": "0f0f4da4d87fc9b0fa5fe514cb7625e34133abd694b73bc0e6234e22393da7c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/score.json": "05dedbde153831fbd10bc2a18a44b1b868fd326e137e81e16bee936587c63e89",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/metrics.json": "d8d45aaf6663d52f18354dcc2265f5071881f583b317550b8ea83f247b12d89c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/facts.json": "29bceaf6669493d06e89f619c8dc811f82b156e2b8e6b355595f750997abaade",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T09-07-11-868Z-v844/codex-inline-judge.json": "dea37d929713c042bcc0bd2cea8f99c86f92cced9ff01ff4c7a4c0cabf3b8a81",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/result.json": "eeef0b82b860e16fd2bd12880d97c77ca149168cc477086d9c73c940a0fa8cd9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/score.json": "dd44b7b706f207aa2cb9004b25f4699cd3c52cc65dc9a63d0259eb0ee2fff6ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/metrics.json": "f8c8ae61235441b860cf17a09e98e42235f200185b47545d04af2862e7ecfaa7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/facts.json": "714b1bfab663271694bb05b8bbb389c3f498f2dcca3217cbb56bb0ad49ca7cb9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T09-22-53-307Z-g28v/codex-inline-judge.json": "fd750d78d589b091e643cbec036cf1377b90d155e1b3e8101929ab50b4bbbdf0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/result.json": "fa8dae639a2a6d85036dc3830e3b27fc3f5164caf2388b23e5e02e72dbe13291",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/score.json": "afee0487fcd56cec43126c11b678955b087bd88495c3aac9b25defee334fcb84",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/metrics.json": "629c47d6171290b2eebbf3854021fc0c488767afa55b643b073e200c56887d32",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/facts.json": "98072db74d4c4380c1cc25c730ced4e791297996663ad44f0dfb811c26222de1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T09-28-26-981Z-uu4c/codex-inline-judge.json": "03c88358ec705f5ece39650967ffd10ee48cd2f3ca1bcaf76bfd671f1c73feef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/result.json": "454b3fad0d2822e4252479b896a9754844ca3d83b789906495bce8eb0e44bfc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/score.json": "460c23e5ad2627790fdef2d6bd8a735f5194740f55929ac3375d101b8280f403",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/metrics.json": "d545a87f640f0736fad7325eec734c5c27fcecb664709d6f2cbb29deb338bef1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/facts.json": "25373faa8401a076b389896da54bdf08874d651cfd41c0d44562ff1e6b1e7ac8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T09-50-52-326Z-qk3b/codex-inline-judge.json": "3c662ac68036c8a42aff9f171dc6555ed26b8b37d1d90024652f768891b3f9db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/result.json": "66d398ad188d286c016d3e6032b9977a6249c455dc01ac3881e44795c34cb7ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/score.json": "d8b8434bd515efb7420107cbecca6e7d76785722024ab4666de117c77a3097f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/metrics.json": "d4ccfeb41f104469bfbbb8a984a30dc8ed17eeb2d529b19f4ee7b77207a8213e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/facts.json": "743546d8147f375427ebe67c72b8ee34440bc4c1fe335e7a434f75bcf107a019",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T10-19-36-090Z-u201/codex-inline-judge.json": "d2543296c8b20e343631321625bf5622721c88e6fd89d138c270aff49cc012d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/result.json": "ed1c4f326924ef914de22ae26f2fcad68d72a13b5537acc8245bd888ae5404c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/score.json": "77bb38c3030ad3edb5272523c9f48b00590e41202a3a0c640d1f67b4ac66957e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/metrics.json": "4087a43494896c92214609a27c959df6ba4fa4f409fa00980fcef5b8774b185c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/facts.json": "04c17bfe1dd809ebd152f3a10bf054630309cc4fc2f1fa8c51f1f8562fa36e16",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T10-20-17-233Z-m6ev/codex-inline-judge.json": "87566da7885bd6d7243097dcaadcd3817627b7568ff8a872496e824d7b2d0d51",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/result.json": "91a7ef94bcb79bbd5c02f237af554c74956cd8bd1e22f46ca39b5dfcf4888a37",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/score.json": "6b9de39ed9f605c8c79598e96453b50901e0d09833dc2e1f5dd0dce10884a5b4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/metrics.json": "cffef744a577b50fc9a92bcf27fc43ea56d56c18e88b1520ae066b04228ce5e9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/facts.json": "469516d1d33ed206aa1a14faf4536a5a66ae455060e454b1a553c15bfda0a811",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T10-38-31-749Z-tcut/codex-inline-judge.json": "7cd6f6c7a0e0b5d5c452f4c983a98d62343c4c54dc3efed3a2a354ed78f40409",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/result.json": "540daaa445233fbc8d8c1d9f28894f0d19d20d88ffef8832c4fdc701d9b5e72d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/score.json": "6ae7aa25e780080eb4eb555db8b1e61e04b0fb21ef34c167c7dd08c2576c82ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/metrics.json": "5e2eb7fa99c34362780f072c7e8980dc6e4f71d1e1f25ba2f013cc4691d711ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/facts.json": "8883318a9b0a44fb751e179c0b6f300468349bee45668832b4153e3a34268aa7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T10-44-05-833Z-g1mz/codex-inline-judge.json": "a06a954dc9d1291a9b8837b5537b74069dc46c4eec262110b5a6aaba1015d259",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/result.json": "e5562b99f909b6aae0a054dcdd96c811140fe3b392ff04c8def05cb177c9a154",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/score.json": "325418c5dcc9a3a3ac885baa3ed3d6d8051e24296db2b50e182ba6be14ff6090",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/metrics.json": "6bb6bd6a470bea620b6b59b6fc56ddd3a6ac43d2856efb4435967e2429e61058",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/facts.json": "8d1e3f8616e94feabdf4ee52ecd90f5fbb9765500612272904eb6edad1d43052",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T10-54-01-911Z-g3v5/codex-inline-judge.json": "75e17a66c495ea3f9a17ae376240ee3b730515bbfec80f07e64a7299d5d19ecd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/result.json": "43ae519c07f2464bbe372d0f0277c45e414a117e8855a4cdd74c14d45d67ee81",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/score.json": "0f0b9b198e42d6fb682dc6a15cd4cbc349d009ac74afbea5664a196e269a2f3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/metrics.json": "a1a1a8bada8b44b92f7682ac8acfa206317a983069037751705efa2f14e12a79",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/facts.json": "8fca2d742909f80d6c51d1331252ef07aeebf6baaa35360cf7fbf6412f23bf64",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T11-07-51-995Z-aymr/codex-inline-judge.json": "f9d6d5f7ecf098ea6518ec120f404fb04bf429c905666faa68c5ce691c97d68d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/result.json": "268fa358872e5e9c973a0c1fcfe681336a9de2cf76d57e250e60b645386b0cca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/score.json": "7609a29b71aafceca63145faad6bf4909692ce17fafed5ac7ed861dd35e09315",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/metrics.json": "bceab76f2b1eab8109acfeec3acca3a8aa4c344ef75736874229538b792ed587",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/facts.json": "37bc216d25aaef4d575b52bfc33ed3485b0c46c803460c70e7b750e34aeb0669",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T11-16-21-605Z-0t26/codex-inline-judge.json": "9966d24af4cb9eb6774aba3682a97ef37176311c4c695d399dba905a608d4ec6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/result.json": "608133a6057dc7c4784bb6cb2912a31703df7f4a444617ed03170f6cc6a5f212",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/score.json": "7a1aaffad678beaeccc5106fb59824535e51c6f3d56923772a713553a7ae19da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/metrics.json": "a0ea413e750d271bae72d3a290d96ada7d108ef35b290e63899c53deea6e164a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/facts.json": "7939f33ef76a2d869fe4db7e0a1b92f2fe3227f307e0c232f5df4b3e4023b06b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T11-43-08-254Z-mxss/codex-inline-judge.json": "e98ea1453770c6b8b0d8303843672a4fd052204bddaa62f7cfe75bc811eeb393",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/result.json": "2477a74e487c565794ad8534e60f44f68a821659f1c4ee4547ebc9a0f076a919",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/score.json": "ef62d0544265137ed03df64753d582386a2c2a326d48adb0d0e91f1c4842d0f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/metrics.json": "657694305efb36ab1d432ebeb1d4a4d9e2c4b40da26c04c1c358f0421adfca30",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/facts.json": "159c3a7f3fd9d3ada682e803b27a4e36c8df6f59bc4a60c748f69c74db89d053",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r1-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T12-22-38-615Z-4egc/codex-inline-judge.json": "4fb7981602f245a3ea20311e91ffd7f1fd65dd172413ea7d199b64a36331ec58",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/result.json": "91ff9c9ffd0fc944243ecbaa33888859d9f9514f681273e4e7ec951d72e77a49",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/score.json": "9861be0e8c2ededccba4c4fbf7bd683f4cd01429fb3069f469e047d3c4d29b3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/metrics.json": "c71bc4d72555ad4c8c5d4c802cf37451288dc6a91f5ebde035f16bc01e2f0b12",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/facts.json": "d7ac843b5fe8f581f3f3dfeb3fb5d2fc168791e63cacb3693c95447c7cdd393d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T12-36-44-608Z-3a6o/codex-inline-judge.json": "a186739dcd2979dbe85c3e43a378e6ce2d5063dbce367948e8a2690d41f2d5b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/result.json": "774e62d464ab868114c27187e0b38124ccaa276eac56fe335fe01a14fdd12fdd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/score.json": "57d257ec1eb8bea31b7350fc6adde488fe2a420cd7a21b38dbc75efaaae97fcc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/metrics.json": "4794a58b058b25ed87bb30f5e3998add3f9faed386df48471a30c8e81c858836",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/facts.json": "5f3d72d34ce4a05433a5bd5458324ef7abe38d169f005e366ea0efc1f852aa90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T12-39-23-096Z-yeef/codex-inline-judge.json": "6d1f0ef5e081187293d64b0bb5f490b706e90e0aa93b5b778a787169001681ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/result.json": "abcc1f86bd60210d95d4f2da02b4169c81343c8a1a5bb64d3991b9d878991a10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/score.json": "dac54afb9ec78102b2c860927a8f248b08577f84b2ca6f38b7bc559d374a015b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/metrics.json": "45dff66c7900da41d933a6b5ea9cbef4d366c7c07adcea03a4bfe9d2c769bdf4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/facts.json": "7bacfa3c473dd66b3c0c11ef009e100025e89679b2bc276aa4645ba4aa1fb4da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T12-43-14-220Z-qyao/codex-inline-judge.json": "9ece89f7d7c8f97104f3b7837ded5006c96e6c10fc4b9d28f1b2f8880510bb13",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/result.json": "35d2a659c2fe505a772927611bf4ff208b923e0f3743e413221538b98c6cc128",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/score.json": "ab510879d6af5c8daa945c3266c58ddf6c0bcfcfecee1302f24978cc8e9857f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/metrics.json": "ce1239299c5d007f4eda9f456a6d7a89d473959a7fe3799f8e105094599c89e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/facts.json": "396701e10a7fa8ce4d1d72d128d96c18b9b8bcb5dbe004d01c2edc5a5389e0a0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T12-47-43-363Z-ehfl/codex-inline-judge.json": "189cb25c76628f062b92d41ebd8fc1f14c1bfbfe96bfbfe6f60c92fc88becfa2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/result.json": "000027116d0d6ba3ec7803018ac41c1726bd8148b9ee0f25fe48942673f755a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/score.json": "83613b55357016201687ee25ff41cc497dfabf1078407c81690b4e2ee7ddf95b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/metrics.json": "6c548716616f57f09bd07ff1910e244fe8ba62ad34ba816c84f883b9470d2108",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/facts.json": "6769c5076b1c4fd87431df7a1b0cf388757a3550e027b5bf8b2ac4f979b63a26",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T12-51-01-209Z-t9bb/codex-inline-judge.json": "5e8a90ec66f1ba0da94c9b64ee8b2875941112b29f1f373353219d9cb321c92f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/result.json": "898d5900af124c3cbfe6c2aa3e8dfb3ed09a29383582ad8c23b1cb46bdd3b28b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/score.json": "f7027d6d7f0c642fe0135db6187f7c46f30f55dc7e6d25bad7cdd36c19802dea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/metrics.json": "497364d3d3a00cd2180f1c7857b4b5f19b636915b6da96fb35b82c0fbc1a0600",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/facts.json": "1f53e5f0c139c8649e56505f82ef5b56c160f4de7f49844b52b4ea88a55c069c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T12-57-28-267Z-js6o/codex-inline-judge.json": "e56d024d96086187a9a45b7488f5dcf1c108efcdda4d456797aebd9a02b4b23c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/result.json": "a8c0ffb14b6cd6ca907dc6aff04c97ec0aa56c49369427d1960ce68e70cbae61",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/score.json": "f99c9cb1706df0091c43e1c46802756af5120067e0060da6c37249b32ef60da5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/metrics.json": "9dbe7eccf10d264ca4730732ca14a22843cfe989b4e6a13f39af51768de014c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/facts.json": "bc68a8be691fea3e9174b574689683b4b81d57f2e87c2fdc4abf5b1639587ba9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T13-03-55-725Z-sp9r/codex-inline-judge.json": "4698dac0c6a8294437f92eaab4cc226017c91210656822c7912190a24f5e8933",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/result.json": "4921393353b2d1efc1f08b9319ac2ce721f0b74ec9bc777367abd89a31d7f78f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/score.json": "0806b9a369aca40876e872150bc440087cc35e7583a1c7e8277131eed2340177",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/metrics.json": "49c5a81ef2716a061f5659df4455ce0bd84fca61bd67bafa88a0ed0be22e32d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/facts.json": "b76f84d5c8cfb555afc35a55e3e1cf4d9ed73e8aa8da3f56d03d77918c45e067",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-16T13-20-05-903Z-v98m/codex-inline-judge.json": "36fd1b784e3fb8530e83ba937353c0640104802953e7ce5c02b4c0663a773c48",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/result.json": "03777f72421c13c6cfbc3f8a627bc2ddc832292d20ef8a1368238d82bfcb0f2f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/score.json": "db7444b67572271b42f79d9a852702c96291264310ef202ea89b4c174664833f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/metrics.json": "733f3afda91de7f85995bf06685619752c48b28573a06716947d67d1a4591ea3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/facts.json": "fd1ad32f2dc32871be5f5ad6b5eaa74600ebc3d74ae1ba4f6d565117c525b3c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-16T13-27-32-532Z-8s7u/codex-inline-judge.json": "dce99ba5b47f697d98e56faea020dbe1aa3603e9961a75fb335c67d099484a97",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/result.json": "d82910c6a752d8ed6af4e70cb7ea592835be939f16f62cf9db25b62191b5d9d3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/score.json": "641b7433dd5911b9609e9e862beee0e82ea12378c7e339ac622d58b18527a581",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/metrics.json": "98ca4a4c4d3a40e20d70a3b550aa166dced0649aeeb241b58c35f00bf8126109",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/facts.json": "2fea56986580411d8b8bbd9a0f8a3687017425f5435d260b1faaec30d4aa108c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-28-29-682Z-lzrl/codex-inline-judge.json": "a0951db3fc5a4a8791df20f56291528a0a1ed38b8a2636a423947dbdc11077a3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/result.json": "2883cf30b723c73e2c4b721714e390e350e8f926d19470d86eac710dceeacf17",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/score.json": "59df0180eb046070831a92bbf9c928c0d508037ec64363a5355457292b2ccd65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/metrics.json": "8718f0a551ad4eb836efaf6b63d8e9edd0900afccfe620e67beab765f93bc8b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/facts.json": "79511716f0822aff91e42f59cc830f101bcf368ccb511abe463496457fd085ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T13-29-36-581Z-ze4v/codex-inline-judge.json": "d1234b5b7c65f65e5ada09b728521a4042f7bde20dd7f7ded96773ea8da4311d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/result.json": "859bff27a6257d46c6013cc35e0f19825f884d66e4db552495f0ef7f4a122c92",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/score.json": "ea19c32e88bf9ba1351b26d7b5bff0b0f155105781a1b8051e0576cf246c93da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/metrics.json": "3707f1b7f99cb09487733da640e9d01140b8b8bbc004f7485fbcb77e4d4591eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/facts.json": "00e549982cf7d4c69183ef767e804c729a51f33ddfa3355736c97261b77a3836",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-16T13-38-48-519Z-ac3l/codex-inline-judge.json": "c6024d264132f942a1d0ddb1573d10af9f4395d5f79b4dffab97fa4aea686e47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/result.json": "8f279cc25f9e4c8be878409989cea2c1cc731e6cab987b8d8d736d4b8fe12f7d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/score.json": "e9bdabedccd4dceac59d297037837210165828c97148b1b23e354b7e44c74920",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/metrics.json": "d98ceee46eef228e0a1ee0a8d92d052b91e7d90d49ed633cc2c920ad15b3f232",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/facts.json": "c808348305adbba58ede91f76b936b7a24b59193efd55eb4705f25dc4d7df049",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-16T13-40-43-049Z-aua0/codex-inline-judge.json": "9964e8d31fed40a96a747fec995b45967952525f59a7c85a6d4c543b315b2ba3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/result.json": "01df3495fc8aaea63722ddefbe556c81e429548f089a8ebe0f6ea25c3da05e4a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/score.json": "92e1e7b2ca358271ff9df9e75754f84da4637b0d31adb062f21ffedce0ddda54",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/metrics.json": "f06585e888e7016a78607271bb773dbb97da115fdc2b36f1462884bc6b67e7f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/facts.json": "5fe4075337f6b0402f632de784ab3bc7ae77750a9c6f10b7ffb7033bfdc9a38d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-16T13-53-41-309Z-l7jl/codex-inline-judge.json": "77daafdd225f9c35b0ff7697d13f99229c99d93c63b0a7a092ce20d9b97f0314",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/result.json": "b7c3a6db0b001c96c44746691277306e22a732348a3bc47ec517d9af5e2c3e24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/score.json": "56ac7cdf38b11cfc0ebb55e70af0d6e37eaf62d3d40147642b32704f2940c0de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/metrics.json": "e5d815a11d9ef615c47f3589b2fe698472b5a1e83e14db1338b7d240c9a406b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/facts.json": "d61abec69db05992b00e6d0ad287f06bfc56e8533930ea44c3c8f945a5bb02ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-16T13-56-06-598Z-e9zz/codex-inline-judge.json": "77daf53d45376ef1cfee59d1a712265518ffd2638b312405bd3b8e943377a413",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/result.json": "29a0c14a9c6237bf45ffe00ac132b8b5bc527a05d2e02d0422ac6a99e17b7d39",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/score.json": "b9aff22bb7263ccd76c969316ae470c9cade8f499debbf9a43fff586f07c27c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/metrics.json": "ca554d6b29b0e7ab55802e7edfe69945799c83b2972d0d3d13b4b4a1ba2c39f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/facts.json": "95bf14f42863df135121bba3610fa11fbfbc34f24e869a2eb170c6065d38fd49",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-16T13-58-42-726Z-y79i/codex-inline-judge.json": "2a57697477f311b02baa63e76c0f21c842ead1d497b57b347b17c358dad1f809",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/result.json": "feb3a597dce8f93bf57cbd5bc92a8d726330d6bd09905e4ec77b1d3b00bfa3dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/score.json": "491bfcfdb0cf76e86dc82b3cf6fb28a81919ca5daf4e84f35f8828254cb85454",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/metrics.json": "ee9454d83cd516ad8c443996b61a608382d8208aad348b7375b430b1c6285461",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/facts.json": "faaa9d242d4393d9b938b70e561a7d4f6fe6ae7b8b4bb92aed9bf163e2721657",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-16T14-03-39-104Z-cj79/codex-inline-judge.json": "6b091b54d3ed0dd033f17bf1da3a559be53c1e199e5e9dc15da35a77976c0b1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/result.json": "77f1ee1161bdd6990840e8f150866a9dab18ce46ee897a1f1e2137a36840557a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/score.json": "30623e2cc12bc056573d4fa865c83bd13d0a44fb5998e933147c230da6e26bb3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/metrics.json": "f7f992549235edb343a4f8cffbb8991411b4e21a40a9481a3437d293b0472256",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/facts.json": "e970642db636cbec7a8b71a9e64838e3570dc04b172a0aecfef091d8e7bf0131",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-16T14-05-56-119Z-4w6v/codex-inline-judge.json": "9f7f662e6bb39c5df6627bfbb4bc55eff7321c00654d0aebdb6df3c17e9061ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/result.json": "c0592c17fcb0b8dcbd7d587604cd3b6966c59421e3165a60ee533af6abdd568c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/score.json": "0edcae539392a504e297e8fedf9e61a182ded1ec926d81874e5df38b1ce77fe1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/metrics.json": "06e8f55d7a23c49640c8953236e77974d864a65bbf6de03145d575f17ef9c82c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/facts.json": "94fbbceb0a6741fd1797138ffa5ce3a235575d4136c89f87e31f3e95744c229e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-16T14-07-34-729Z-9c0x/codex-inline-judge.json": "e742e1d3b98d4d9ac86adb05cbc239818ede7e72932e9657629d80063620a208",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/result.json": "e745e221ccbb2cc9fcbcc67ee5b04bee650217b5ca822afe638a0d6bcc60bee4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/score.json": "4ecebd3248d7875d2e2d2c1700f528fccb41ef247e2c65a59e18b999e958af92",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/metrics.json": "d34878b8a3501f89bee1f3db016a46331f47525aa1622b21b3ef0a0e972f1ef4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/facts.json": "d06cc5b1fe0264cdf4263559481f6d073eea7383580a01385e91ac683966352a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-16T14-29-10-347Z-ig3y/codex-inline-judge.json": "61b5062d05b630c27997204d6a0f725d0479735a322e5ee09ddc2a4a95bdf244",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/result.json": "038d15fd55659f9274444e05eb6134cba4464c479d2c67f26d0c11dabb0ee22d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/score.json": "3bace8e8577d604fcf19a88e5578b8f9d9aeba18e5fc7800534a741be082efb5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/metrics.json": "d0e0112c6c74fab2d15f2d91592d0fbd6d221a1808b72f08a89c58c5309ca88b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/facts.json": "005513c1323e1117b42a6ee6f7a579a232d748487a4d910524ba4e067885632e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-16T14-35-25-053Z-561j/codex-inline-judge.json": "8496e509c3806d617d82e9c40932f3a0d34189871e1866d60d56337364d5335d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/result.json": "3cd6fef01f3efbce669fae59a319cf9855d94eb9f49da704e4124a594f770ee8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/score.json": "72850cca2016c75cb563570d5de90c7518caa0d4c9d3180040f49818dcbc6fbd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/metrics.json": "4d92728df73b4d10e6d205da2455cf427ee2beb0cfc5d6ead733f8f455b7db76",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/facts.json": "e4540acce7c81add750d6f73df8e32895a9f8e0a18d21295f194b31176996370",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-16T15-16-11-704Z-ojgt/codex-inline-judge.json": "0c84835536de15dc9c10e252c709259d59072e199a875a8f849afe9859b3249a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/result.json": "4dc098f7ceab2976a32717157431c240f774e619f93b4489419c5c2f61a67541",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/score.json": "60021d25a0a7972b4c5fdfab2f96fb5f3711adb5197df039017773ca4ad7fe54",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/metrics.json": "e6ec1ee5740ad7d0018723d9491130d538a20c3af93f620e31f4bf7db1015cd5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/facts.json": "a441f35681ad0389fb309180d3379f18929fccae824545da93418b3a88830720",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-16T15-27-40-506Z-9xj2/codex-inline-judge.json": "712d90caef7d586a9e69992f1d40587b45d61e26b2aad92196874fb9a7449678",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/result.json": "f9724b3604ee486db3402066581f96f487d0ec8205c7a0ae4512de9754976f93",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/score.json": "08db7c56620ca1fd6327e8610fdb1b359448b4960e39f7ee22619f3cc322daf5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/metrics.json": "abcee52d80e1d8f6b0e5b502d183403172f785b818ba7bbd6a99badf7b5f7316",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/facts.json": "147a24dfe19e637a771309815256d6a08a830d4f66b99500e7923af49c764626",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-16T15-33-03-869Z-e3wl/codex-inline-judge.json": "e7deaa5f6f9ff98c5354769e0bce20d05df624b5ad5919dcacb84511348df215",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/result.json": "95691dea591d2d58a8a3fe65c68bf8d395460071be6d8fafe435d7b974da2157",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/score.json": "d6b1f198041132d9c002dce5a05eecf004da54aa19199c22f620c7d3faa261b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/metrics.json": "8fb6904cf0101633bf8b5d399eefd7c8d17686a993a939fe0ae7aa7d30688add",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/facts.json": "8367131f0b9c93697b0fa7b0ac2ddc14a8c54f06f8e9394c97499a859fe9bd75",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-16T15-36-16-085Z-unbs/codex-inline-judge.json": "32d5c2b29a93d219fbd16ed33129bb722cda5b058da2ea2f27ea8acd678a044b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/result.json": "8adae259d172b5d47115a780ef32ebd7cd6e01e7141fbd000a3f4379b143ffbe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/score.json": "d8c33af0587804329ed25596e5f427a3e15d148f2324d4ebc452c4e5ad92ef01",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/metrics.json": "573a0d74b85ec330539472e3a47f798049e0d1249da6e18f6191f57e5e1cc10b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/facts.json": "454fcddfc6ca147f0cd97d8721e67d5928eb6fc695b797e373e1edea67829c20",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-16T15-49-00-678Z-y6hg/codex-inline-judge.json": "81e33de7bddb4113343e83dc9b7c33d25fc9f2f1cce4775a443a6858c1d0652e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/result.json": "ba5113bd75098deba228adfd0382a28de2afb20852bcc697be304242b8a7b23f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/score.json": "74538d4620708e44a401ad7dc75c691d573c7e1a2b47df57dfc270ef721d3797",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/metrics.json": "290581081a7a54f525872a20112cfe9cacd68a858d7795cb7175c9e6989f43cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/facts.json": "3ae11797f261cc2eb28304111c4ea6022ecc47c851f8d177bb36ce51bf36b8d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-16T15-49-48-884Z-qrqj/codex-inline-judge.json": "1bb5cc3113c1fcb02b703a8f43d676f148c69721fc80d34268214de54a99d120",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/result.json": "560d89f6dfd522271019febf0f72e09b40f59117eeef593f1c27857f25b3edf2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/score.json": "9e6b51e3a4e78d274fde85c3099b56fbfbe3c9b1c83325c5ac4b8a077ac78074",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/metrics.json": "7e2cc6b6aea24118977f68f07b8393e1c5a66bb33c5194bfaf3f6f5740f1fdd5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/facts.json": "7bf3ef0010e34f52eb7451e1eb8e9b7a78e9c170136eb043a2c8833d6bcb39bf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-16T15-52-50-084Z-jo1x/codex-inline-judge.json": "b48cc586d8475c1dfceb7c351b308458599b1e7768a82c90cec2f25117ef935f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/result.json": "25f36c8c8d547eb4daf1180f8738cca416dbc4c0fe1882540fdae2b7163bf1ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/score.json": "fc0f7038c229d94bcc349cf5facb5308a56b36668cb5efe1679961fc2b37658d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/metrics.json": "98e3c87fcf58b4fd0900b7f60bfa5786995488c6764c9c9e5d28136a102ee666",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/facts.json": "2899f0bcc68cfe919719918247c5fbec6bac76dde5d899464b6f0a0f8a5544bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-16T16-03-31-944Z-qyuh/codex-inline-judge.json": "4880f3d9cbf6053de814fccf3f0121d3d56b4fa2b228362c849affe8e8b139e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/result.json": "452ccf03d7b01cb4deeb08fc062b6e42dfd8a4e961bf40ed9b4ab1e2ce6ee779",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/score.json": "165630fefc72686e5feb23af13438395ba3aaf9e86949aa370a91609eb6dff58",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/metrics.json": "a594b4729d1cf77703a7b2f8fe56aadbc7e91b5c8ed1e2267eb48195aad641ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/facts.json": "e965af4c06c8c1715617e5a729e21b025f813b52710b1a05105f1a3d4aee204f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-16T16-13-24-554Z-yr4n/codex-inline-judge.json": "c124029f394f54e4f81a9819f7ae39707f18c8ebe20450f4f29d5029b57413ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/result.json": "a64e515e6899ab6c56dc6a941489bd79f03b31fb0193596dd867b29a4ad0010f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/score.json": "b7495b95829b3d6cfccb17e302c4d1832708835727ea09547b48993b1729dee2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/metrics.json": "56995a841ed910060e09fb860a36b6e72e06bd8a664f0049dbc9c41ad258b4ca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/facts.json": "272814d6b7d1e482ff5f3250a06625928154ad4e6eddb671a584ef94a04c7c29",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-16T16-19-30-725Z-havr/codex-inline-judge.json": "8695984ec3e18971fb3b511e3f15a3514224d9cea1ba6006ede50aed88f5d88b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/result.json": "17d1c6b06f0970a60dfed59850808c486364f1ce917cfe26b2e05277af60e749",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/score.json": "3dfdecef8569099b1c8c4b452c3869fea1c200cf5460e4b3d4d44f7eee00a865",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/metrics.json": "d97dbfb24f8ca9891b0bedab7f60ad4c097e350eac723c6b542f5c920a28c4b4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/facts.json": "fa00fb7c455064196dced4e2ad5d3cc4dfc14035c4a2d5892e805797eb3470a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-16T16-36-17-072Z-9qjq/codex-inline-judge.json": "7acffcb24c5f31a7c5055761242058687a6ed7d5e85a2107e856823efeb79352",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/result.json": "05e70f7610362d4fbd82fdef6ea4b17d77584312f78c59f687e6126c6bc3e313",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/score.json": "77f8b4734385fb1aae51ef7bdf6d026201dfbe9893896d2bb56471f4a612edc0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/metrics.json": "e1b623bb75e02df6efb9e37b03b32ee37005c94115ce3438e03b8befff32fcde",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/facts.json": "015f1ee2abb834848d1fc83d022bc231b8b8237906c873c7329ae7b30843727b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-16T16-50-28-009Z-0ult/codex-inline-judge.json": "b4b31113ad2b1f9facb442ca8b4bcbdc75b095389b03b2364b922242cce1b542",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/result.json": "1d892d10b2cb55d03c1a68e1918475bd26538694bac8a516b32c81ec00f3a50e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/score.json": "086343fc5f81d345994e0686c788df28c0f4d655007bf0fee766a15d57086006",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/metrics.json": "a72d49a466bdc435c9b06ebe520a0db11a4fa16b409ac8c7b8ff198ae230f9ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/facts.json": "ad489d7c107f44e843cdf5a431e0afe0e2890a39c463c190ba5e3800670fee3b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-16T18-07-01-700Z-k9hy/codex-inline-judge.json": "ec0bfe30e8d36bec68de116af9cc338c940172878a2b383008437681f5c35d59",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/result.json": "40c361ceaf090cbb1982beb4a73404f84bc991dfe8aa88cf03522a5fedfdfb77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/score.json": "1363053edc07f279c5ee14928a273fb5e2c5086479bee1f1d90c2cc0ccb13fc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/metrics.json": "0fa85593a9f7ca6d0fe372452e5232844f68c209c2dd3bcd58ec905930f8a483",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/facts.json": "5abca09c0585f372510f672e7c550ceda01b3d04c083a990a6cacc228098cee0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-16T18-24-20-717Z-gien/codex-inline-judge.json": "c2ffdcf444b6599d4786d2737e91e9f5aceab78427ba60ff188fcf0a8300dc65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/result.json": "1ba4622fb873c3a85ff871a10c7d9f7d0a6d15008b47cdc6f2c8bdd45be5ba4e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/score.json": "4c8984819e33676f576c56b93b274a1af77f7932a56ac9ed4c0a66b2fe232eee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/metrics.json": "85a6e3728a6a2fd675c513b8b9005cdc39d87bfe10a82d0a1c479f3eab8b59e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/facts.json": "a63c8caed0fbb88668e7eda539eb6b79760c533b0fba71e4fdcf70dba1ccb471",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-16T18-25-07-077Z-yk6j/codex-inline-judge.json": "5bf096cc3015cfb829265f3ceb95f85cf319db76000bc620be8d3deb34a364be",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/result.json": "c14c61f0ff44f7f9c745c33bf7ddd58aa28a0000e3a11b9be7d0056420b9b6b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/score.json": "408b1f6951db94c6a10999eee271a0b5571d4248ebb320eaaba8870d69e9d9e4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/metrics.json": "62739e846bfdfca3f975dc32c4e78c7efedd2bd22ebddde7f6b880533408d274",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/facts.json": "86f2219d02fdbc3d1c3db9f7cfe15e65b4e0689de0b347e58c405a5c729d7f51",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-16T19-30-18-651Z-vdnz/codex-inline-judge.json": "a5dba9d69e17813d90d9ce795c8a11a41e825a54b50faece724c6ef49f0ef16e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/result.json": "0e29f2d11f27c68b74ddd1d5dd0d00d505b985726537a13b099c1e7effc0039c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/score.json": "2c0ea1963e4215dff1888844be69a6466339f54a7cc6820f96123b344e2c971c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/metrics.json": "dccc1a63af2af36ee969b326a587932f8a5523e082e69479b3c70e4a58d7bc86",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/facts.json": "7fe817d15b7507dc3a774029b26b1fc216f55a7c15f314e0d93539747b591dce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-16T19-37-44-030Z-wj48/codex-inline-judge.json": "680066ddd276393b6a407d02833eb095464e10dfc3d4868776c20af5a0d6ab47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/result.json": "fd027d9b471b86a9aa69b7fcf6b16d58c5ef284061236e2b337ddf4caafab5b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/score.json": "60ebbe22ac7cbf4f9f281185bee22d7ef9cf593f027076533a43eb429e9cd089",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/metrics.json": "26121d809e82457c10fc4753c99b73245196199187745bba2a138cde093f2cc9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/facts.json": "84956d8b167180701f31c6f73f1b51c49ca1a8305aabbfba82279eb44d885216",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-16T20-12-18-990Z-zdex/codex-inline-judge.json": "bf45ab2593b9eeb3ed62e7fe4c87e2ebfc6feef1be7c02b5ca19c625c4775540",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/result.json": "4dabd8f68af0de9422c3c87a091cad2f794e43c29a2de6c93d5f03740f2e76d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/score.json": "247f5e1c9f42e51d3f875d162c76647893ee9a06e57165b1161a6363acca2930",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/metrics.json": "a6a7cfef6a45161a5eb8a013b0bf8f3ea9303ebb43a596db489a2e9aded75a88",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/facts.json": "53883ced8ba2d84dcb2a6549c859663e16d8a0d7a4df9912144cdd85e81b66d6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-16T20-33-39-844Z-1fav/codex-inline-judge.json": "eff3e7128a4c83a0de054dbcc34fd3c97075119ad0a39f6a2497ef5251563c4d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/result.json": "343d721d5c309f8671fa7763d05f4fbed13bbd57165b5ba61de04b35f939aae4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/score.json": "8a3d6e310036c6c37d8f3064d1242de07026deacf2891f19c52c4bafd4a0c667",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/metrics.json": "8f64ccbf307a18d83d530f0055b20dce5a686a4215884ab7cf11f4ce142c95ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/facts.json": "c66f98548fcbd5b785a9c3a09ea6222457bd6332a16504bfdff3330f1db9bc89",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-16T20-52-54-796Z-488k/codex-inline-judge.json": "caaaf825c50502302985882c71990f62be5065c506b315014c62b74487be58ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/result.json": "4e16723ff03b5c1f3b23b2f3ebe63a73b5a0847f5f776148adcd7e294c89f6ab",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/score.json": "555f948b9aa83fd5503889fbd4e48fb91d5d2d7098d9314fc26dc4db3d188a59",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/metrics.json": "3db50b4f421fe2880a917cb656a02e44a38dacdd967f44358be90de8e8764912",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/facts.json": "a732ced6b4fec52b05c2d6ce55d83b994a62cd599e2f3dc574ea14a14503a892",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-16T21-22-28-031Z-jv9g/codex-inline-judge.json": "7fcc3b6646932e829a6efbc257bb7380b7b35a2daf618693b6bc698086387fd0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/result.json": "bbbc180339c878fb94be6760efc5ff06afbba35ead52defb375631c5124e6b71",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/score.json": "55ce2a039a2f1102cc93fde972b6becba9a200c9e30cf6b45efa2fdd0e8d29e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/metrics.json": "6b48e0d0d5b0f8ba5c02eef4581b2e7561ae84380826dd64a6ef0c1f6e07857c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/facts.json": "7b8f2a4685a76c994a63091b944bc1f1d2118d43359962f83cceddb8da4a2455",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r2-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-16T21-59-16-125Z-wulq/codex-inline-judge.json": "2068ee67210815e5823f2bcaf58b7443696aaae0424c64f5dc040e297487293a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/result.json": "1b8746d68d85fc5f6af70922f89e8a17fe203b32df0c856d4c8b0968a14dfdc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/score.json": "34db206718c9bd79307e47f2d2c55b25382f9c33202acb200c3c50e1a25cd30d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/metrics.json": "62db12352191a7f514ca3af72b338351e715d742137ceaaebd8e4e0cc429a22a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/facts.json": "ad4b637d375384c31daea0d44ba3ab5c77273577675c97020f80c37961fc0be3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tictactoe/tictactoe/tictactoe-lfm2-5-24b-a2b-q4-2026-10-16T22-04-01-260Z-wofx/codex-inline-judge.json": "3401e9f24e45ba99dd53928307152dce21a6174b75720cdb84303d6d79e034e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/result.json": "73c0fc8831a6b15da4dd7ec7340eedd61487a87016b9808629680c81f283a6dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/score.json": "81ec037ced72a3a1e8d802d5a0ccd0dd7f24749c93598c20b2ff9797b5ced0f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/metrics.json": "92151482c9b65f39c92e74d63f05f4805410d9f79c36e6fa3bd6331ec6e7efe5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/facts.json": "6ef538eae5bb706ebeb9ed54eaff8d85c4b88da049555fda286167e04137d0de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-petshop/petshop/petshop-lfm2-5-24b-a2b-q4-2026-10-16T22-07-16-708Z-glnt/codex-inline-judge.json": "7f8ee6d8a7b346e566e7b6bc68578c6ba79477d8bd1d107d48cdbc84479aa743",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/result.json": "ae5d5cc817ebd9c08c4db1bf257b28dc036a34db05cabb61ce6f9acd01125d78",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/score.json": "5d6a818b89efad331c7e1f582c3cbe91294bbf29f4f8a26cf27586e219491832",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/metrics.json": "42776f1de81292ac745bc8438de1a41c5291f1b3bf36ac5a4727e25e8234bf7e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/facts.json": "afc0e32fb50689859729918150100b2d68aecce49fcbf31ef2b04f523723c8e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-tankcombat/tankcombat/tankcombat-lfm2-5-24b-a2b-q4-2026-10-16T22-09-42-571Z-opyz/codex-inline-judge.json": "8e3455cf8f78385f8d72e735c692564ffec2789bc7353df966e3a0483ed86479",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/result.json": "99c176783b554bcabff52fed21aefd163b62f52ca11e099356b49c20d16027a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/score.json": "752cc62eb35c234fa02e650b010cdfe1b2e2e1db81bc3102fe87925bacb2c3f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/metrics.json": "77ca3680e0c58da2fd73bbe77253e0c72975eca71afd917ece4089c15e606fe4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/facts.json": "171accaa7c7a1650902ff52efb728c9a72963a70437313550f16f2d7ad3d8bd0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-schema-migration/schema-migration/schema-migration-lfm2-5-24b-a2b-q4-2026-10-16T22-18-55-077Z-dtki/codex-inline-judge.json": "5280aaaf7490cd2b8f209b8d74fe44a561d3492f34d83c4d3165cdc47be53077",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/result.json": "cd9b6f52ba3f6da3846e490dc3629e1e5a25709f4d4756146866c8d3334a28dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/score.json": "8dce68fbc0e6e533061d2131ffcedfd28fa43b9ba131a3af2f7574e1d9a1500f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/metrics.json": "1559cf0acb8ccd3c3e62c01aae59ed4f304a704dd01caac7674f1a88ba8a366f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/facts.json": "c55c1d1204d0ec2aaacef6ceab6dc1f0953faff45df4449aad8caf8f390da822",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-lfm2-5-24b-a2b-q4-2026-10-16T22-23-45-757Z-xow6/codex-inline-judge.json": "131df46941748c45b0bbf180b1ae774b50b928e253667ec25b0f96f3d867e001",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/result.json": "ef32bdf948890d35e9be887a77521bc3ef17af528568078c6b51467ee6a51295",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/score.json": "01dd86dcd23559c336fcb0880db8ec331322afec0ea373262f61240f6425e225",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/metrics.json": "a298603f77c7220cec84f0f3136f7c0466ff1ab3a41d9cb34a75f18482dec197",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/facts.json": "4fbc67f647ef9f9ee7fdc966f0e35452ea35c26f3a0d272c51286ae12f021ce5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-symptom-debug/symptom-debug/symptom-debug-lfm2-5-24b-a2b-q4-2026-10-16T22-33-36-164Z-wzm6/codex-inline-judge.json": "019fa4eb39d412e9d1f5683bade935bf1f3da0787990467dd7b36c0e145e5cff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/result.json": "6343691bd5f16f954dcb202e1f9d6996d392c7aacf115fed01fad14c6c47afc5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/score.json": "47f76620e0ae4029958ebcff7cce12a46607b183fad2d4d3485cf5cb91d4de41",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/metrics.json": "dbeec097c42fffad1d4a6013003d70fe6c77d61004eeb36ff0ffd924e103f716",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/facts.json": "201d13a1179beb3e75a77752dbcec85bc7c1da534a37281d5646ebf52d472bbf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-data-wrangle/data-wrangle/data-wrangle-lfm2-5-24b-a2b-q4-2026-10-16T22-34-52-726Z-zx7w/codex-inline-judge.json": "761fc2f264f95f7663f96f1235e3c04f772dd01bce8141099241dcb368c3770a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/result.json": "9f4ccf871eb3bc5c1c24ed9b61dbde67cec8f85472f10aa17b49efb281f10bd1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/score.json": "f315eb3be925184ded575b00146c53ad39815183d5eee7c319a5d59a054e452e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/metrics.json": "f5cd5abd242f0740466a7f07bcfe0be86c4210afdc8a1da71823b6d1754a2910",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/facts.json": "403ce82b42c5a9fb52e7f57c6b011c2cdc0dd634e2170d516bd9a536812bdada",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-lfm2-5-24b-a2b-q4-2026-10-17T01-03-31-777Z-a5nj/codex-inline-judge.json": "e68a46b69b2b386bf02fa356f5d5a9a1bbb463957666aab5272219b06cb9c553",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/result.json": "1a43314f9f8b72aa623bf857bbd948248fb66c83045efd8617826e481e312b32",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/score.json": "191e73590f7123f74343543e141a53ecefae748b8d1276c49bc4cb29470543fa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/metrics.json": "2bcac0fa79bc98b9ceec80db976e8c063fee23b2cf35934f937d64879be6613a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/facts.json": "dc43fdf680d48bc2d6376add133e2f015108b4589e95b734bd53227798a726e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-lfm2-5-24b-a2b-q4-2026-10-17T01-08-15-398Z-b3oj/codex-inline-judge.json": "b4033747682e79521ca5237b9a6b30e935dfadd758b618ede5798c67e0cd46a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/result.json": "96f2e29c32c6f203a1a1584074754c19a95471f4684b3e63f8febe25b155f968",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/score.json": "bb079e56ca0d9e3cf2ed5c9e8b7dba10b914c801978b821e6c783386eb0b9f27",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/metrics.json": "8af7913b41332c19f367d7428a76e149d79094d6c786fabb9fd6f58fc917d61a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/facts.json": "f9a895d54bdcd554afa50aa821308a5ef2587467ff1c4927229cc12b8a36b81e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-09-02-574Z-1kp8/codex-inline-judge.json": "9f335738706fe617f4605cd367a598229ffaee67e3c121364f2516f02666779f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/result.json": "0da2168481b3c92cbee0981e4388e6c3b7fcc5147125c92345bc849ea70677f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/score.json": "22d25c8eee7e81bc389d5805296d6de1de19e6639ef5220f74efa4ad3a743b8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/metrics.json": "94da5f585b12519f62d315bab0b6cba6ea1407c1c4626a1495d45e6bdeecfc2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/facts.json": "479dbd555e5a5cd613bd49af5c6e035e10d246bb1a7fc5d81806ed6eb390d5ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-10-10-131Z-48uc/codex-inline-judge.json": "d571f316b6f70c01cee20959aeed9da2f45d543efc7e0b57fa1704c461a5ee8c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/result.json": "2d7b023c45acf4f04992a782febf98e7192490bbd7fec24805fde8357aae6728",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/score.json": "d10489c86d326eb2b7e34873600766cf77ac34483b5c1a574a29f8b5a5b948c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/metrics.json": "e23f5e1de1da5213c2d4433813c2998519a759f7cb15764efaf7ba97a21c9aaa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/facts.json": "94bf783e0a53b78a6a3f2e2977b5d4290013d8e30ef5f46f805c15a3da43c425",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-lfm2-5-24b-a2b-q4-2026-10-17T01-16-54-947Z-g1hs/codex-inline-judge.json": "0ef6057a3da9790a337f342bb4770c6678f8414c5e805f497d6d6cd18a23530c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/result.json": "7d0179ce8d8e25641ee1826ca6da1241a5b212b54cb37e258b48265d574d8874",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/score.json": "7be2497a0846f966461ae97c2d27e40135a873c5741f7beeefb5f84731f898f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/metrics.json": "638139e8a3addddc0587e56f2153f39e4192595f52210ef37ba5039493b91913",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/facts.json": "6c411798c7bca91e1e2c277ff16f5868fa3c8421c6f3a170bb4a236bd3abeaa4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-records-intake/records-intake/records-intake-lfm2-5-24b-a2b-q4-2026-10-17T01-17-30-947Z-kxl2/codex-inline-judge.json": "a1dcc29095ce88f02bbe2063c8e4bceeabf823afed9d10ee0133eb10a04e2fc4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/result.json": "383f8ae521eddd120c4f080d0ef6a9599d770f60ba381bab50565410d4038517",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/score.json": "09f482ed8ddea12b9829254a5f4f5074d58a37030913924151f7e14b53e770ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/metrics.json": "478e75dafe1040c8162964bd815e78a46580b6e562b863e67920636444229d56",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/facts.json": "0f1f276eec8b607dfe5f3f0172c586a429c0a611f92a51fe2cc26c3c899b7b94",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-lfm2-5-24b-a2b-q4-2026-10-17T01-20-29-069Z-nxa5/codex-inline-judge.json": "dd64526c15342a1e25c52e9f4808c7807091c3b2cf04b91d66e0e803119f2044",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/result.json": "06902ab85d87776c1424d24c94704de66bebe3274b9caa3c363a82b253a41362",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/score.json": "61ba577928cbf5f9d97cec1df47d309bcfa9bbb0f474b0ab2f143732e5a30ec5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/metrics.json": "acb8368a3bf180234fbe41abb59b060f0565ecf721b15875e3297a9a83bd045e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/facts.json": "75fce7d8ed1bb51fb2963bc2b5ff06bc01c511095bec310f3785e9c366712d65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-lfm2-5-24b-a2b-q4-2026-10-17T01-21-56-513Z-dx1s/codex-inline-judge.json": "0b80276a7b2bf8004cc8233112cab5a54285ae665d3b5accbdbd0ea94d3b5cf9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/result.json": "3981eb163e7bafad36a381668ed849d6bce3a77f3bb80de7fc13a49e28d86747",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/score.json": "16a6f44ef59eaf0861263a88c2cb0ee9cafcd4b3ffacfca549272fe53e4d874b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/metrics.json": "d53966b19b6ab98dec9f27f55d507d71c54b2786b68354efe75b068c1485f26f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/facts.json": "e50c351a44add1d53d0e4d93b2a3423d3951edd0bc2bc5d6b379f1355faefbc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-lfm2-5-24b-a2b-q4-2026-10-17T01-24-13-269Z-be94/codex-inline-judge.json": "857f75b9b098ea84335f294db6e0adc248746b57956ebd98feb97e17f36509d5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/result.json": "4302e4295569857ca2b8197f6d4f8490aebdbdd03eea81a083cd6dcd71a3c1ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/score.json": "15ab91e0ba69a2ff8461a35dd1f8851b1147678930a6a0ce1cad97354d0c479d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/metrics.json": "b70f612aad36aba94f9e96b2fed4f2a4bb9823d09f9453b166d9572c20affd06",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/facts.json": "cef8a404b332195500db54279ed6e4735bbb27ba9b9b0087f87ad254a1369269",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-lfm2-5-24b-a2b-q4-2026-10-17T01-37-06-978Z-ilbq/codex-inline-judge.json": "97c5b2dea2653c434ab36a636b0c89801b13b3aa54e028ad77bfdfb862d9ca50",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/result.json": "20b281f95fac3badfab4a403929b8ffc9a3d227b985565f92daaf3c5d09cb9d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/score.json": "f307e7352742cd037b642d10ac40fd34cc3dfd22f4bcf6ab4d8b9cd9b62e8893",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/metrics.json": "7476d810c70dcdb575435f30e4a213cee38dd42fe72dd1c4ed40a4eeed265e36",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/facts.json": "ffcd0016ba38ffaa50af2887b4f4d3cf2a82e544a7da79e8dd8017990cc88fb1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-lfm2-5-24b-a2b-q4-2026-10-17T01-43-05-502Z-hxn0/codex-inline-judge.json": "d0c1b4b06b332d2869a9f571bab855062927a1e97168c9e0a120e22cf87d7b94",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/result.json": "be11f791b759c9924d0c43f31ef943d06790a9a96bbdeadf924d37dd4bc32d99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/score.json": "fa0ff1b55a4528801a3e7c85c28a26f1e442764a12f5aad1e7bb99d1b0838643",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/metrics.json": "2390f7781403311538c105d7be5477810082776e31c7e5b89be4c88c78cdea6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/facts.json": "e9c08b50d407af6fc05b714a54b2f0efd83f0ba546626f12fa3773481d61cb46",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-lfm2-5-24b-a2b-q4-2026-10-17T01-45-22-338Z-puu1/codex-inline-judge.json": "ec0b7b02e387a37dca1b460433972b673696d1826faeda76ec682ad6a5a009f2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/result.json": "1a9f9ea8e749c63d1cf03da93e364be06879adbbe53ca7d6f37162959b75b100",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/score.json": "601d01b0adc55a3ddb856d1248c95fa3f5ba14b0d0507e627a4904cd6b3a152b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/metrics.json": "8d23d786f7d87287ff9cc2b9e1ac9f99e8fdcf87ad67d4b369380065e0a728cb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/facts.json": "9aa68dd646980ad1311c66b4b4a490158569c3119451e5f84905e6a6e6a2a626",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-lfm2-5-24b-a2b-q4-2026-10-17T01-57-31-398Z-tg6f/codex-inline-judge.json": "414cdf37126926400b701a65f976106d486263a1882ed038c186cb0fa7f8b280",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/result.json": "f2753b7bebeb2ed9df8eba2f93d868c647e4db9ce47e55f3dd98e90a23b0dde4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/score.json": "f129fee86bafa9f09a31bac8b946e741bebd47b8b9cc0b492522eca9b5d3f02c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/metrics.json": "408fab49d95103176a44d70cec2ce119b0c452594a335d80ed17479d562534ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/facts.json": "83b42983dbdcd2b7365211592a4bf9863b4a10ca31f5b6a8586f5591d9603303",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-lfm2-5-24b-a2b-q4-2026-10-17T02-05-32-036Z-lmus/codex-inline-judge.json": "0b7ba93a2b2a8b732b230ff18f7dc1e1381ff456ebb5199d21e1b9e3c5b6b540",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/result.json": "6b30f8de7e6f2377f31176b3b81a6cd4eb00264ee4e918d3f96892645512c069",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/score.json": "962823c5af119cf9014d3bfcb2b78e44710a587b10e7fb8cb8052afb36a6eb24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/metrics.json": "2b88177b2539ca33da01deba4173a82eb11510b6c8adda1dddb653e8b3c9b86f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/facts.json": "91916fc3065ad3dc34fac5931fd8c7ceb4d2e03104c8b713c75d838a600812b3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-lfm2-5-24b-a2b-q4-2026-10-17T03-20-03-865Z-i8zj/codex-inline-judge.json": "dfaf56344729de7b7a52a5aae5e409c2097652592f561fbb3b9dce8adce1a4cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/result.json": "a5e10528b06e60d1bf56be55a72ab4cde3751d04fa07b388dede83f90c3a7b68",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/score.json": "3c38eb99a717824022d192387d54dbfff45e125b9b1d8a64d4731f1199ef02f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/metrics.json": "91f6a0b33c82282eb300ba4ad95c3f168ee657a97a788dd602268f15bc3b3208",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/facts.json": "a60f16122125d5b32e43a9bc617a26e6b95d517d265a9d6f125ff2dd960073e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-lfm2-5-24b-a2b-q4-2026-10-17T03-26-48-668Z-v1rx/codex-inline-judge.json": "b668f88b3c1df47079c157952744642ad763a9f8f79e29ff865c23d1b428994e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/result.json": "1b8f4d82f184b3226b65d52041682fe6dd6c6536d5490c215137157c75f6f925",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/score.json": "b55930b3aaa17a81e73b9c8a01a58863d940edbee57f359ebe2bc591549dcf65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/metrics.json": "bb187c7311d9d918901bd84bf1efafc5cb702e215dc36e4c8ffd1ad41eeba2e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/facts.json": "4369f0f83924c6125108b74ec692352d2c3893d4a52a75d127c6a9c032304999",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-lfm2-5-24b-a2b-q4-2026-10-17T03-30-53-204Z-i0d5/codex-inline-judge.json": "aa7504eed2ab7a1779c867fafb4c73723c9ab5fa781379d280375a94ad488987",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/result.json": "79ca3e2930e97d8a06c9e3a72651d8f8683a071c60a8cf24df2236601b627559",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/score.json": "28f3d09d12db3586eb9953ca8dbe8c100e25c66e81d9695419adca7c7849c282",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/metrics.json": "812b5d3f65ad533b5611e5ed541b15ef1594967dc20f3653c04da49c40a8ea34",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/facts.json": "20f6b47bd4e914a04f1f9396114a03cf85d147896ab692c7936f495823a21a4d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-lfm2-5-24b-a2b-q4-2026-10-17T03-35-22-920Z-ezs8/codex-inline-judge.json": "719704d05898fb0830829f1ed4b3a5ac89127341e0db30fffec3549e724046c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/result.json": "e479e133daacbfde606fa360345f68731a2602c6186ccc94b1a0197f6217744c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/score.json": "b655d382496968aaaece118829498a48e4e4a2425f311b3e2bd9dedc06deea2d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/metrics.json": "937a54f3d4b89a1dee5161e5eb1e9fe97ba28c35f73196089be01c9926568259",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/facts.json": "9ba483f2cf2fe5684099a36f8552ce10004cef4d3f0007180066e9dfb3a49351",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-lfm2-5-24b-a2b-q4-2026-10-17T03-49-05-821Z-l5e2/codex-inline-judge.json": "9b062839149613edb230399d4f67b5ba659e6ac8544b65bc2893849edb9daff5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/result.json": "a2c68b21fc34ba8325b47db5ba6dc6c2ecd92d45a8f545bdd991ce70b95a6056",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/score.json": "bc62f322b4c2345d52e4cf91f63d5e474916dc393273112a010389f94c04f3ca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/metrics.json": "eb31bc6bf319545c3ddfc40f180c832a746be21aaebfd6f92065d6fa6b7ecfcc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/facts.json": "8488d12d28a6e14762001c20e5a7b1ffb28b825f412f5ce1c2bb00d0d0dbc3e4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-interface-contract/interface-contract/interface-contract-lfm2-5-24b-a2b-q4-2026-10-17T03-49-36-759Z-08yy/codex-inline-judge.json": "8380e33262143a40ea4d4f66ceabb91869cd5d38429c2b4fdc3722a9a126a304",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/result.json": "ee51bed1f70c5a46fab2bd3321667485afd122950bf403f35574343738a365f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/score.json": "1c8c887cb97af2ce5fb89ac73d70968967ed840a0327189af1d4cfd8a10ea46f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/metrics.json": "8926d24d843eae06a2310e013be7bca8f6791b8f36ad05105e5c5c91c11970fe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/facts.json": "84d7284f32dfbd52a363ddaa3f377f834c7e11ba35ee863ec814e07b31741596",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-lfm2-5-24b-a2b-q4-2026-10-17T03-56-40-633Z-i1ny/codex-inline-judge.json": "7f417348dc1c2c7dac9f9d0390483ed8d344beda7cb39e9ef877653b1fdaeb1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/result.json": "ce4a149aeec954dbe3fd4fa4fe519016ed3145695041cef088eec43d4309ab19",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/score.json": "1f0d25f02a51036e7f639960a98f9bbadce7623d33672ee2a23d6796634df9c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/metrics.json": "8f5ff990274e137aabdc20666931a27b0cde385f70769896266800356f7ff1ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/facts.json": "0e50e5a6b3544c731adbe89fe85713048d1fb1d52e8fcb16f8bc1ab535355a7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-lfm2-5-24b-a2b-q4-2026-10-17T04-01-19-170Z-uv9u/codex-inline-judge.json": "3a684aae70aa1bfa9dbf25205eda424e06cafb104647525a0703b461d874d503",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/result.json": "e1807f21ad14fda037e327a617aebcb4c0c3688748154be864acdb21705f0feb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/score.json": "9a011b3eaf3e94d110b13a35ea9d90e36269a130066394de582fff553fd10e1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/metrics.json": "c4673d6bf80e53a2c5a6b0b35f713d4693f5e461a914ae74bd03e56eeb5cbf3b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/facts.json": "1e1fa99d7dd42669c131402cb816328f5f768b3e67042caad3179de0e2c546e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-lfm2-5-24b-a2b-q4-2026-10-17T04-13-01-459Z-o63x/codex-inline-judge.json": "8fe897b37b4fc06188ac9a489268a187715f6d6aabc2c28773bf4c3679b8fddb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/result.json": "4da8050fc126aca95685723d60098500ecc0c57859ed1126b808ef0eb3196dfa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/score.json": "1065a533f863961a0e0f7aded6b4ee0b127308fc115781eef04fa16fc2237f0e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/metrics.json": "604c063c7c015b6a2add88db93b766d0f2cbfa17e5d8ac95320f16971591d783",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/facts.json": "2f66f36cbb9ed3c08a1f2d68cc79b7bc8bd2c78ff43b4bcebcb1dcc151fa78f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-large-pr-review/large-pr-review/large-pr-review-lfm2-5-24b-a2b-q4-2026-10-17T04-18-24-396Z-w69r/codex-inline-judge.json": "f5d987f200214cc28a16800e655cb189fdb57aa82f8832bd78b48e48d2029a28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/result.json": "085db523e9d575b8605b7919868934eeaf7943bb3034725b50348ae8abbd3177",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/score.json": "f3da6b530852abdd6a1769e88f3e7f8182eb717052d9de31ea115b9b66861b0c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/metrics.json": "aa1fc62ffaf0136b8b9bfd755d359da5ee5c3c229dca7393630720bc0372e5de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/facts.json": "a097d54f12451524a76fb231f760b78f6186d7993492d21cbe8b07f694491685",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-lfm2-5-24b-a2b-q4-2026-10-17T05-03-31-850Z-42jq/codex-inline-judge.json": "1c9899f647ec092126ebda970dea153bdf1a20d368bc973385c9c97e2e115e7c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/result.json": "6da2453c84c8687f2add8ff578c53609e690a6a5087c04f0969827a7c687b1ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/score.json": "dac9f4af313cc73b68c57714aca27426e594dc0f7cf69759355fdc56ac85bf8e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/metrics.json": "59a7a2fa00171a899d25dfe5daf722cae4bf9569522c3a809046171f9a78b735",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/facts.json": "87b5c1089cad5a334e87f1d24a7baa0d118e4cfe805b1a59d13fcc0f886982a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-lfm2-5-24b-a2b-q4-2026-10-17T05-12-57-475Z-qqbk/codex-inline-judge.json": "022fd6b36ece24b9d229ea3781200b6f558ccb7be2f7bb192a971a08f0cb92ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/result.json": "30a65123d1c289422f37779a01f1878145c43069c4cdbf5fd4fdb9e550b2b1f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/score.json": "bf984c86ba7907ad66f5a8e469db60debdbf304c2aa7f459bcdf5da85749ab27",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/metrics.json": "691d0c0c44cd0a8084a64e409f2fd9a2cc09bc1b5f84e5c05a323f4a75e445c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/facts.json": "1cb39ae1fd4dd66fb36cebd66a06197a832f9130950f4ad958a74811c342a9a0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-lfm2-5-24b-a2b-q4-2026-10-17T05-25-02-510Z-brjs/codex-inline-judge.json": "1a158a90614098249e32a0fa548364cd11143cd0e82a7a0a4db75eec25b01e77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/result.json": "33e10192a8558004683bd77b893a048711b9a953eb0068535665707bffe38129",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/score.json": "5aaa57c804b308b72727e05e87472bc7754c06872aa9ded85eea352baba274eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/metrics.json": "995a48b36fc250bed46d4169363b68876150be3be80778c2ca9a99ea31729984",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/facts.json": "d2fe7dab02075ce1404c7e4714f0061c3c51a109b2f5f2f63a0e9f620d3b7c45",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-lfm2-5-24b-a2b-q4-2026-10-17T05-39-57-205Z-5ydf/codex-inline-judge.json": "34acceff7129e2b6754f670455a991248692ea318cb17d008bf618d1b6aaf1c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/result.json": "d6231435d2cc0d4b34df20cf1b057d8249ef4947d06f21186e0ba6862fe29b99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/score.json": "cd6816f794584e8575743f69618a0dba08c622235e6514615a57a183f91f8f36",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/metrics.json": "8f3c630a56d6a34b011b86d3e3e771681942954b685452a998664d037c639048",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/facts.json": "6875b1564163140e3b843375e5685ce33de5f2aa224db83ea13acf6033663ad7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-lfm2-5-24b-a2b-q4-2026-10-17T05-40-37-224Z-qrdi/codex-inline-judge.json": "6724827d907172a09b9d5b07ddafca478e44a35c972efdc507e387e09ee001cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/result.json": "1105bfe0e5409cf2e20db9a721da7bea708212de9b140361675effc8c84547fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/score.json": "92ddc18486df52a1e7c30b588ef1d0eb1629433c236508b36971435b30d63774",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/metrics.json": "7a0aab344073db296a4bf30cac8413401e7ceca988d4466f04fd969cdfdb2c00",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/facts.json": "3957052e789ad8d5f220f6c9ed78de03ddeae245113b8345395def671dcca970",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-lfm2-5-24b-a2b-q4-2026-10-17T05-57-22-033Z-4wfw/codex-inline-judge.json": "b8cf1ccfc119c9d8e129f4acdb4b4ef55f300450cd318c66c568f251e2db9bd6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/result.json": "bbdaddf9ed479a580f2cd4f63b9a8025e1d11f7b233029ca1d6a5342e5eb54c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/score.json": "c7f396f4552082f8a50a3efd45f97fe66a7caf2c04036212e8d5ea5fa1fa09cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/metrics.json": "11c5613bac3a4b0bb39daeb65d0a70d4fcf2abb41bf0998c3da7275d294b7566",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/facts.json": "25cc54d8236214251ab7fc86e10aa7487735a3728721d6530a1213905b1e6147",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-lfm2-5-24b-a2b-q4-2026-10-17T06-03-35-504Z-hb3o/codex-inline-judge.json": "41bc39e8415413b154a5cae6eed63d7e962f12ca6a9f59e3bdcc9c41c31e21f4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/result.json": "bb16bcad4eeb512096cbe90560d50db388841e9aa004460c9221afccde236544",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/score.json": "7c1c499811d4990c2fca68c10e046d70e9c59256dabce40d4b41fc4e8095f7a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/metrics.json": "352de0d18c723010f61cb1b152f22462d8f6621964606acb15db1b87ff6b1ccd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/facts.json": "b484a4377f0b73336f0ee2577758770730dd0c8efe4f95d6f6f3408161531df3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-lfm2-5-24b-a2b-q4-2026-10-17T06-17-36-581Z-jedq/codex-inline-judge.json": "af9bf13e4d1221ac7a293ab2a10b7748f0e798e71157f25861a8f79aa992c804",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/result.json": "2ad8d0068b56bd1316acf0e6d037d57a9fc3c551726f0345b6bef404af4eae00",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/score.json": "03b56bd250e3b6d114a674c869b64b4fa11420c14ba16379ef7c2a0364429f38",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/metrics.json": "ef0c99d760b9bd46d86537ba7e998b38e0da93745e63e30516442ed213cb9e83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/facts.json": "0187714e718b99b88b4ed49967f95f081462c31846655e158723d6018f6537a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-lfm2-5-24b-a2b-q4-2026-10-17T06-41-10-627Z-y86n/codex-inline-judge.json": "976cfa6964fce8f4b5367e3e41756b4e0a28ddbb00b20e0dc9475ffe5d891e51",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/result.json": "6718df0fb75e2a537c764d118ced5a3d30667577241d5b8a297aef90941ded75",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/score.json": "c512cfc7496d414aa828c03dfba8bab101664600e6f739e124adf004147d4087",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/metrics.json": "69451b511fa136fd4e83ec794bc9aa0d95eb0368ac4bade0a336dd919f241b0c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/facts.json": "05f815bea995c04d6173b204ab734a27ac775ad3863ee5bc361901520d9b44d3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-lfm2-5-24b-a2b-q4-2026-10-17T07-01-36-834Z-bd70/codex-inline-judge.json": "6700bfc5f90aea57e01fe5d39e10aea0900e129960d756d22bd70dd258cbb1af",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/result.json": "e3a26161baf0623e3d63c42d46e489853dfa21e5c53f9d1e6080abc4aaf36af6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/score.json": "c839c53ec5b6d1183801affc5aa25e5b56b4be43c52b1f97b61c05dbe717ad95",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/metrics.json": "e3bef8e0a7415a06dc078acb76895970665047ad20158c217cfe28f4037e13ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/facts.json": "15b0afb419173bfb14d88c3f5b6d3596fc4a3b9da334dd8c89ed860d5001b87f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-lfm2-5-24b-a2b-q4-2026-10-17T07-17-08-699Z-uhys/codex-inline-judge.json": "e2b6e197c8b894f79d903df5eb43ee3efa94f635c594811e499519895ac4c0b3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/result.json": "2565c24e4f4afa38aa9139237dca4e86ef4a2d653317b1a0cafa9b4a5c9e4c46",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/score.json": "deb2973731a650a9cbbd6f22b34ccb30935c7d72af5e4406f6c872ce4daaa10f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/metrics.json": "2301fe99832c03114fe9756ea9863efeb1b9baacdc10fda110e59fc753818c12",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/facts.json": "246aedf819d2d54081ee26dd0e8df0585446f68fa11155e119af5efe9396e935",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/lfm2.5-24b-a2b-q4/r3-lfm2.5-24b-a2b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-lfm2-5-24b-a2b-q4-2026-10-17T07-38-30-467Z-y0vb/codex-inline-judge.json": "e6edf2cf02f44da487a7cb46a96d208b8784a55f163dde9446ea927fbe3b0aa3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/result.json": "b53b8cb93d2cb42747c14e5b848eff014dd60023a12c0136f88ffc8b673e9165",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/score.json": "6d5d46d3fc6ed72b17805b7d2c153fdb1bf00e09b14a95f00dbe048efc6a6b73",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/metrics.json": "aedf1ea3511be87d27db69b64685070ac4fd67613921e4180113a83a8baf7872",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/facts.json": "46a266a7296e02f34b2c7fd28e06f85e7d74281437f2d18d507226b901dfcd77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-17T08-23-24-461Z-v0ke/codex-inline-judge.json": "b99a3907e1293fe37de1f92bc4a9835c55c1790b8d2eb054c793503f805c56b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/result.json": "1c0b054d277d4cfc80933f9ed0f53e7cb1d59881f821cf99488daba61622a1f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/score.json": "698035155b8b8d08d03172907f6496084950abda813337c4a9993b267c0c8bb2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/metrics.json": "f11c5b177336b253461b078fc69c34f86297b6a7cadea1e97c37a9cc97f4924a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/facts.json": "0ad4a82ef64165e04c46621f60f037b952f5e8b37ef709fe4e30238ab8bc246f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-17T08-24-44-682Z-5p6s/codex-inline-judge.json": "8a2005f3b7ef6e18e54155123d6a8d9a56e17ee33ef21521353795d265ac2945",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/result.json": "efa5ae315f5cfa6716e8a52c92a56d100e741786cb2fcd715de119cb94e3d08e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/score.json": "5e0542ee77d091cfa8bdb59e7afb9aae988070e597599326934454c3a865fb99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/metrics.json": "6a5d28896aeb3277dbb97e1f8b87abaa01486e970d0415342f2564594a8107eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/facts.json": "2c77df9fe155ffa42534797c7316729a8602c25b6e88a0718d27f10d8d7359d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-17T08-26-50-195Z-6hao/codex-inline-judge.json": "3a52f8e17d106c19877769b416ee81b8630692c2772f40dd79bd99ccd36a3e04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/result.json": "f47d890dc91282bf39e07be29acf8e85524a3c4aad01984a9223d6335706dbac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/score.json": "b55e361c1add8fa6addf925452ed0df7b832f28aed4622dcdd3949eb5db2b301",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/metrics.json": "408d29a5a43cd298a10fd911e059332e7613cfc54d8e28a63c41e9468b1627f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/facts.json": "bb1d0f794b33bfec3db8a6dc90e80108a662f989fecbad868b578258a0c58cbb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-17T08-28-16-382Z-bvtk/codex-inline-judge.json": "cbf14b6a737beb067d6ab7d6d2327b3f12be57c6064a7d4f8899206421154c3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/result.json": "b5fa06288845363ba2c60457816a9de8f0dbef4be8688bb8c5a7ba0543cdd4ab",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/score.json": "58a557b15bc33adc454192ce336d23b88ebee7af1eae2dcf47652603e8a4dfc1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/metrics.json": "9890f3ee75e14185461043461e0b43f921423397d9e5723df0d04d55e9fd021c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/facts.json": "ad013d0b1179fe9a32233c1fab6df83fc8687285f347d272064faec7b827b9da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-17T08-29-43-802Z-bv97/codex-inline-judge.json": "9130cb2d932083a9d32c4c46039e77bac798a54cddf21b6c64df75b7cc4d9dbc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/result.json": "e20716455a954ee7eb97c179edbeacda28674abfd30b758bd5eefa1488d70839",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/score.json": "9ea9c4d5ebae77827e51d4c5aebd81be88fa43f0d8875c6ff282f4df95c5f6b0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/metrics.json": "67d4378a309a16b7ff177603dab9c97b7a60749ac2e89c9168d8910913033e7d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/facts.json": "00fb8967e099fe2fbcfe673ff9621fb898e7670e5828ca849701ac6e75b4a320",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-17T08-30-15-773Z-t82j/codex-inline-judge.json": "f8c8e67a5b479d007f969d49db7e17f82b78602d60de4ae4ae45688738237af0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/result.json": "c4a9335b76230ba038d91454cadf7a29ba976d6faf226a1edd2a267ed4775239",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/score.json": "fab0794b8e4247a1954c6ab0af280ba914bef3348f25909382516ba224c6f9c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/metrics.json": "2c7164a3285c12594637e67f3d79815683f87b7eebbb1d80aa2cc74bae3b8fa1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/facts.json": "8724258c624ea03fe06aa6a6327118d8622f92dca4fa9d948877662e094d86c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-17T08-31-11-854Z-0ik5/codex-inline-judge.json": "e87648ce3d5038ae7da2dd0e8753c9acc41307d71e3d5f06621e0247e28b9a6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/result.json": "16041b854d7dc1adebceec6ab38aa6dc0661e9f0c3eaa8f97c2f03ab5d8c9bd4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/score.json": "624d5cb36f8bbfc40b8d786e271e5bdb1ecc6732f1e9eb838bf73074d4695e2a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/metrics.json": "6610f446a945eacc4b97dfbc5add7d1ed48a3c910008bcab8ca71ba0ebbb574c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/facts.json": "02b37b848e506d46b174139746845638cde0bc77f0f7651c7326ab826569c5f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-17T08-32-19-006Z-jnwf/codex-inline-judge.json": "1060ecbbbfcb14a97966cb807367a34f094b1408abfc0c447c6958bf65dac255",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/result.json": "703d5229dfe2d71044858e06f76fa16869ad65e2bd081778d5438627c5b4625b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/score.json": "c1cf6681b964db514ef1ce2886a9548587127f01d6c45f1add610b4481fcc9ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/metrics.json": "c1ac0599f4471bd5edce56701d1482e7c0826eac4c62fc795053133948590ac8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/facts.json": "7f39c336b196e47f3d21974108eddfcfceb3c6e5d4d12f4310aff29164925d26",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-17T08-36-26-862Z-jlkv/codex-inline-judge.json": "19d4befe25e6ba5adcaa7f02805e14151f1fce87e1bff7a8898669a3cbbbbf92",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/result.json": "8bad010eb1ff999dbfea9ec25b1c915f480b8a45af3471e49c4894adbe588430",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/score.json": "3cd7ee7d059eefc5ab851b77ee70dfd25e22b5d4d7689d4c2ef259a50f7fc2c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/metrics.json": "5929cd1ac7d7c3a23d9d01d10ca9809176bcfa6182b8e8aa3debcf069480bb69",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/facts.json": "3b779fe3b6c9bd149a3949ebd6cbbe8492ae507cc32bdcb76a104568a42416e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-37-32-899Z-8ruw/codex-inline-judge.json": "6a3245204aa43ac75fb499c16c2c3236e53972c93a71c4d05b092065fb6c2d68",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/result.json": "90af5d33a3c95a6991b8dd618d9ea95fac54556ee1cd570fa2f9a5dcb0558266",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/score.json": "856b9bf8f6d1853af28ce63b39b3fa37d2403f7249c78ea9095a1cc202771017",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/metrics.json": "3c18828ccfaf50ef4d9a367a8d0d1aa60bb849542fbdd9e10502a9d4984ec5e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/facts.json": "8bc252db02ef970bd762851b454fcd51d4cb6a666d6efdcf73d88cbcdfbd3597",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-38-18-870Z-yhck/codex-inline-judge.json": "f3753e5dca38945362b81e0a1e5471da2df81ad411b9482cf962f87d9cf82994",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/result.json": "d427d056b8b715a9049073307f2777c887ffe5240030fc77cd0de0d11fb93cfc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/score.json": "4332d493d4151d414823df0dd65fe60114dc7458621a81200d46881ed0d29d57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/metrics.json": "8791d8bfd8b85fe7214ddd9d9a230279d4717dd3c9fec70d0ab2d5e9923040f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/facts.json": "5a1a42e92d2bb7a28e9855fea50765b5cf8dec96f23997d9608afd207581b374",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-17T08-40-25-234Z-r5th/codex-inline-judge.json": "1fcfb19eb07103e19cc3dbd7349ed3b114f2317f94056ac7228d4b2afd0a70ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/result.json": "e8e0d94d2094ad1836c6aa8ac316e89588cf6fe052029d89354320c863f8c364",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/score.json": "4972427a4283d2a275d07b225deed29f5a9dd47ce62055ff12e79cf26fe719ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/metrics.json": "41db1dea9213c215a06cd2dd8da7d765ff812dd558aa538203ab176bedf581fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/facts.json": "80289bc7a0fbef7e1b481313ee7bc137804f7489097f73224032528f19d1d3cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-17T08-41-16-315Z-qnil/codex-inline-judge.json": "2927b82feda3be5049dbbf4de428f847d805acecb562d7d631224f3cc55d0bc6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/result.json": "b100ca4c3bb6b0fda164b3a045031b62101981eed13db17e8c73e4058bef622f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/score.json": "cbcb635b1498901077ff49b264d2a906a927e56d7f5bf97c923bcab930052f60",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/metrics.json": "a877fe273c292b564036d680f690304e3f8a69eb9bf271a719ee8a117e5fb736",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/facts.json": "a0929a05db57fecc048e051cb16d1824af87e1d2e310846583452235bf85fb2c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-17T08-42-53-506Z-3owg/codex-inline-judge.json": "7de8e9d1ca034726cb574fe8be78347c0b6843b59543b4b55f1aa6b1f39c3eae",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/result.json": "4e6026924d4cd47b6c6d4f2affa4f78ece33e21d68ad778e8c18c19d28cf26d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/score.json": "3a0b00b66cdb46c55375c7b2fe2d522cc1810d1b5b7c6b48adcfe44ab67a0751",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/metrics.json": "bfa763a3a1123739b05025f6f592920c11f0bde500c2b79d60fcc23af35a9f1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/facts.json": "002ffc516e3ee4165d8907e476317485a8e79dfb9cb1790736adfa5f415cdb8e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-17T08-43-29-467Z-4vss/codex-inline-judge.json": "c8cd109d832c023307cf3f31cf04ef6c7afd8922aa1f522bcf2902c264f84337",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/result.json": "0035263615f709645270541c13a77f78df6178c0c18026139b9ef7a91e8b22c4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/score.json": "3d8e18a8006faace7ce9a27e5165b64bbe0e424181f96315c1a002b9b1ae5cf3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/metrics.json": "4793b05ca477b74481dace8ceea36938381053140cac623c40bae88192ac9814",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/facts.json": "f9084e8b7a27ea75cab779a8c1a2889fa7301a3671c983b0ce86b5663e7617a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-17T08-45-37-126Z-z2mn/codex-inline-judge.json": "3408e73b4642c3397eab8dac88eacf9d8f803db89a7a83fb48ef7b8894200015",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/result.json": "0fe19705a607c0c16c5f12462b85a308ee722571ea6c7a7886f563c0844d1b83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/score.json": "f1f64b417643316258ff28b853332a57940e79af8fd5d352d86591766258b5a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/metrics.json": "4ac08ab2959275d6d42c96a70d2ae6f1aa875f41eee5b18a97d7f741afa4a048",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/facts.json": "f3ff25d4ce45d1919594e60270f169c3a45ebaaee03a9bb0b2e041e3005d1671",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-17T08-51-12-367Z-i54c/codex-inline-judge.json": "5a04eb53cb0eac35dee80048bc3f3bd544dba99f487dcde11f7eae561318f0fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/result.json": "dc051b8e4e791d97625ff0c952a6aff680b00069e5dadc6d4a1d13c4d77fc707",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/score.json": "9dfcea855d072bb3cb7514ec20d5ec403bf689114c92d9a306a1de7240d26ed5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/metrics.json": "cff665d1f78784cfd5e52eb7eb42df7d954a91971ee9de6f4ea194996ee3a34b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/facts.json": "35a6ebcaf69fcae79a55a52eafba78bc58f125d397a0c974861dbe0dc20b286c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-17T08-55-05-209Z-tgbg/codex-inline-judge.json": "0f08290ce3226dad552b783f50c54cb547e101fa8a0c58b829283d177219cd58",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/result.json": "8bce095afe6f0e1d5baf1bf191993cf27e248649bc054f1d69eaafc7021b00a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/score.json": "61c105655d4246385560103ff620070192f5be5f4d22b691ecaffbe120f5cb3f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/metrics.json": "b0fb118eff6bd5d22c82282487ce0594194228b16a81fe8258b0722b612c5801",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/facts.json": "01abffa54fef3d3615bd1e2e3de4fa07f5d779e2d5eb536d3e4d3786e4ffd5bf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-17T08-55-51-202Z-75ii/codex-inline-judge.json": "363091e190550879424c33157a37753826732eb7af72512f5896e64bcf13c323",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/result.json": "e0bae21efd8a0ce883895724462972b3891bb583fec067fc74750d0be1daedde",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/score.json": "3fc2318930921a1272bd60113f8ebda5a6e35fd1b6c4bfe0f59b2b9eda43303f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/metrics.json": "7603d21882337b2a5da6261903d2e6d13cbb169ff422e01033c6471cbee3d150",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/facts.json": "903fd04f1854791905f5684ff29c7011db7f1e8352c2e884dc2fd2d11ab82e87",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-17T08-58-38-739Z-0aby/codex-inline-judge.json": "d69103e3efa2f85edd817f3546ef14d5b5940da6027fbf4488c3152cb6b88f1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/result.json": "25eecd039255b64f8a7c31d9aed700f0b348a0f02d6a4b427e0450a9b69137dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/score.json": "69acde8d1cb154e4e782ed8c080cd0dbc390a6ccd49c59cec5e32e8a6179aec4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/metrics.json": "7da6c17e22f99f69c058434fee4feb0ff60b6640196ce472ed3b2b978b15d693",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/facts.json": "03b285fe1815545dc4c295f48e9ebf9d8d0776c962d86b219cb2439bd993cde1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-17T09-01-11-293Z-tbsv/codex-inline-judge.json": "b5eaae73c2a8474f83c276a7f9a1cdc7ffb9b34e38b6768dc5b67a4346b9898d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/result.json": "76453298609e4c1a4afd82144956ee26a83942749461489ee6515a996f1b9687",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/score.json": "cb7f08454f92b72606b6193422cd06eb213365a87bfd61a65c32f1c4fb573a9e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/metrics.json": "43240214fa3fe2fd6a53d27a0646ffb26cab5c8f5dff503cd9ed2782db8c60ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/facts.json": "1eafd2518b0fa32c54655b1b4f77443463a300666f77262f7dff2d4e2d020273",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-17T10-27-55-721Z-bqxf/codex-inline-judge.json": "51f642a2cffb730868f55527839b275a947e58d27925a9ad86e4e32f4ab8a93e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/result.json": "976341f45a483ac23ea8f339eb705ea2cf9efe8a6a4ceb6b1c646d8ad310dece",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/score.json": "a140d1cdf2a49fe84e01144a377ca0c8d5bb01d2cb7ee66b3bba3a685e6f88a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/metrics.json": "1d68cf8a81f7d95d202ae0606ea5c699989c43f35e84606ad656f277c060ed6b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/facts.json": "699216a815c82b36f10d7fbaf04ee29e6ba0cde617599550ae1cbdc538e635ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-17T10-33-19-240Z-ys0w/codex-inline-judge.json": "817c0967c04e5043f1b526ffa05011e019f7b923839c2c555724f145ff742f1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/result.json": "6bd8efdedb4a9884a99238e033772a268eefdd5038703d02e4e8ab059707762e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/score.json": "ec787f1bbe68eff295ae0630f9e2fcef8197a067eaa55f0061e9f40274008809",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/metrics.json": "aaaec1a6c998373467eb3d424be1396979879adda9fec16e77336ba1a278613c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/facts.json": "349f1d8073394365d58012088e9b30bb700bc81a3ece9e3838bd95fff6e3f7e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-17T10-35-22-982Z-ay65/codex-inline-judge.json": "2d001aa4c415bc36e3bb9ac6794a25a86cc6e6d1450d88c08bde1de4c0c8aa28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/result.json": "e1b73ecf5ba11edf93848de05c4b410f1f2917f8511c71c468a750697c56ea7a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/score.json": "65713d10189fc4abfeadcaccfdd591e46cd4a202405252b146eb404741360a3b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/metrics.json": "04e6a9e5df47a5c56670a6824c1b9d722faab22108ad858c36a4683566aa39d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/facts.json": "145f808230a15e4ef75a8d10e5822a1e88aa90cdc7e0a20e92bb2001e7dfc0c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-17T10-37-55-745Z-wdtd/codex-inline-judge.json": "afa81820ce8293f7cd62535227f10415d6176912b154d7ac1c2a891436da69f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/result.json": "b2c1da413c5a51b769555732796ce64739322455e93828f3049cec3b3d34657e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/score.json": "67a17d4e5c3694592a2f3ea2492b88fc9416ee0aa29bf6c39779f69d293e6e29",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/metrics.json": "7f20d6ddf38654e0cca3440e93210d8d023f509a47d3815777123530d8baaa2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/facts.json": "9e5e9e62ce410bc320153d2151074eeaa8164327ad3a6969f9a9ea4a8e84f171",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-17T10-39-02-829Z-ygli/codex-inline-judge.json": "cfbdb676e1b3eb762b92693873957d4769c651bbe9a7311946ec07b8800398ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/result.json": "e396c771be1792d6fab5bb0b5c6e4bfd6cd7a9b9ddb46de71694fd86075733ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/score.json": "45027d06726adb13d1c9ec0f80a858bb07f14083c05c3895fb64bfea947795cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/metrics.json": "c6433e09565996b7786397bf143a7be40c5e0322523161b47c7a4002af8a9790",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/facts.json": "547cb9d50bf099e97876038d868f557de9674f64123cad03662d1c16e67d9752",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-17T10-39-39-758Z-5fod/codex-inline-judge.json": "2358b8386a8665e4f303eb84d3cf2a1638d0bd2524db3f2a39ec8af8a1b58b06",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/result.json": "48ed5bf380667b4f46308c7f59ef56f96c56a81e576d831615f185e6dfc7df30",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/score.json": "85e3125aac65ccdaf32d1b6cef11939187565f619790031155cf0de6ea6dd000",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/metrics.json": "c56adc8fb3c24f5b9fb088787e75e0c823cb685b44844e0478d341d4b6ab99dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/facts.json": "5d3345bdee34e0d817984cfed6bc1a539cabb79d937941ae8a3566d940940392",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-17T10-40-25-724Z-snhh/codex-inline-judge.json": "37fc88bedb46e2812564d546202a23a1509f94cedc0452c4d08e43fe7fd858ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/result.json": "b7da5c67f1504652e698e41c7b96e6cfdce6d0b8ffd57806357a0b3043853fb0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/score.json": "e2ce6372b5e9330588ea56552604d7036574f2108dc00870dce337aed166d36c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/metrics.json": "3628f3551b2753f1a1e9f0794ea27f181bdc7460e653c93f63771c0d60f1712f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/facts.json": "dbc4168457fb8fbb9a62fc234752a5b99b06067a902061ab8d762094f4632e33",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-17T10-42-57-184Z-sjmj/codex-inline-judge.json": "f0dac458d3096582be593362799e7fa9cd2789a8b2ee4432793b7cb8dbe15466",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/result.json": "c6723df6bc4fe752a2b44401331d14a687c53cec1c5c019037aafe129aa58d47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/score.json": "178875cb7e969aad639edf2f488e1188214bb3cf9a13e90d06e9fdc10e29e4d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/metrics.json": "f76d4eb97e0dea2b3e5de46bd481d98389fe22423aeb5b9a40ccbf76a9cbdd4f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/facts.json": "4e84fec8d4e125329aefd1bb2a236f06c88477990b97a37a68467d4dce87b850",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-17T10-49-35-719Z-8sgz/codex-inline-judge.json": "360cf50bd62434b7ac792215cd0a4c1dbb30b20eb435665f85a09b71cdcb15b0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/result.json": "92e4d8659452958695d8893e5c67e779f66d7b44719bb1596466f1b97a191a57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/score.json": "bb06af6ae6ac1470b2e1e10259ec1c82392d5dfef826d509be21cff80493a241",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/metrics.json": "c8f6680b8158a96107b2a2d5bf1c69c2ce857de0ca43e74f4d09c4f19e5490dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/facts.json": "601dea984885312b78dcb74c60d7c796feeba76b81c58a90f52bd5ec5ff655a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-17T10-51-58-141Z-6vvr/codex-inline-judge.json": "d7f03a2d2ef7c481d1248b967aeed367e16f0d80327406fd4ab395747fe51cfb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/result.json": "0bb290f00d359643a9dede7cf5f88daf852a06d72bd8b8466850035430505e7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/score.json": "13f1d6af7fe702b45e87b8b76bc8a23d396c3e00ae41b36101f83d1d72f4c08c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/metrics.json": "944c03afb0231cd0dfb51f5f58a19176a7875cba31c09e970c81fc57f614af00",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/facts.json": "ce644f49455940aaa80606ea3850ed9bc6a2f0f069bbbacdb0572aea74b8c74e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-17T11-02-43-632Z-rfkd/codex-inline-judge.json": "332b45015093f346633f16b28ccbf838fea3cca965b0072801a469f75285733a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/result.json": "aab47a381fd22b95766092edeaf825e42ec88846f408d39879a142b44693d4ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/score.json": "b77f2cc55cd7e4d9dcf4da41eeb4b37608e9fea4f741fcaa5eec77fc98426490",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/metrics.json": "b5e6cfbee2c133d78de0a9094e61289e20f2e34695ced0a7a7a28d4aa7dc50c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/facts.json": "13bf779d731230419ef7df27efdb2e2a5b176440ea9fe8a4ab6acf03831faef7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-17T11-07-33-668Z-w3pe/codex-inline-judge.json": "257c5c3329ebdb21ae4bf177d33034c11bd5f305e18439490fcf6dcb0ddac932",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/result.json": "11cc12e4addeb5f8d01fab00352ae1094f6f69a295f0997134f72dfe8be8d95f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/score.json": "db8b3d5c03b2b167637da401f7779f5549ac5eae75658b369793514f763bc414",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/metrics.json": "f3920aa5131d1e6b0aeee35d49128af530a8dc1fca9ce1410a66989a94dd317e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/facts.json": "d1b6f681cf6b5d7a6689386148e0abe489763980affbaa2a10c73ea228a62ca8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-17T11-24-22-789Z-bxj1/codex-inline-judge.json": "b7f090b20da46ea7956fd80210666c68056ba5eac1743a99c86044722bf76001",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/result.json": "da9d3bd256286f8e00407f71579219549c2c5a66d2c0b6abd29faaa746afba10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/score.json": "c3516ea72c67d9a4131e942961efdef264ecc3b238b45d834540ac3636d6b824",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/metrics.json": "127042d30b7b849523d9f9e74ac1426b3205dc84b3d1bd68d63a632a75f75769",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/facts.json": "a514ddcce590f7bed7c4295e4c4edadb5498711c39e6e454e2f6722d94104888",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-17T11-30-44-238Z-n516/codex-inline-judge.json": "c346c6fcd3f0deab48bf15cdceb6d2f2ff04ca4ca11e60de830a972593345bd3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/result.json": "2b53f6f32b9a5822876b717df85a2239959599abae2c4efd69c6f5d2f58a9656",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/score.json": "f63834ebe5991ee04dfaa44352799164a5ae942e233fa5611ac24eb4a49f3758",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/metrics.json": "ea2a6ee6d36a86480fd39705a63be789c22edb11d366193b7c922bbf4799096e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/facts.json": "8bf751adb4e466adf3d7ecc402274b47fb48e63ec8d13b74258857ab4917ff8c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-17T11-31-25-359Z-qc28/codex-inline-judge.json": "5b564d0562c14c219c37eb63038fecce8e37b9c29c205aaa02f1638c3127fd43",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/result.json": "4d470b01e9a0cd98e369de1dacf365f15b061df869854781f376c363b26c6c04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/score.json": "c59923ed09966bdd676689697871b299e62a41662eaa2989ff14d3e28318e2a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/metrics.json": "d67d972425b4993c76004f0b87d4160157bfa6e166b752502accdfbee8dbfb1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/facts.json": "78f3c0f543eea2cea5264a5b76c1fe7b6530001afc70bff8af4e39547e1c3b56",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-17T11-40-18-460Z-tlo9/codex-inline-judge.json": "629f57f02faa7d5fdc397c60a5eac8f79c1f654d3e604d02fa62dc49066af43a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/result.json": "d1b43da3d27a2eeda2c021579ce7bc1503689ffd3ed2bf15e05ee6bfc86b59cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/score.json": "91289944dd592cf3ce9f6368b511824435290fa93190939b767780b78fa83bb1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/metrics.json": "f782874632c1a23e91138018a956ea4e16c844a3a352c4ff19953b7454c173f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/facts.json": "551fec892658fe0e8e9a6b42a6ff888d6e9c7402743f5bd37e485b79d69510b2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-17T11-44-40-503Z-heih/codex-inline-judge.json": "3bb7b0ae342d88b428831b5b0bb399752759feb12b3ed547f286f6a80363956b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/result.json": "420edc4682144400281442b1911029beaf3683432e8e772fbd263468e752cb70",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/score.json": "da56cc5f66b699a7daaaa0f738dcf99e0fd12435b4572504b23f1fb2416d9448",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/metrics.json": "fe8cd1b9258aa78e50310a5c9e0d1083de260b023695ff34e6ba96546637b8c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/facts.json": "ad8bfa51173accb4b4f6334fd7eb621aa0810cf787eab2c5803c323289270810",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-17T12-05-13-775Z-pnri/codex-inline-judge.json": "d5ebf0b069c434ae6e6224db796b98f740bfe4d5b0ae9b0724e2cf4d7575bcc6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/result.json": "b96299bed4e19a7c055c06c3f88f51303469179d215022e61ae91906065727a0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/score.json": "8515364f2defa822c0014c7a47ac9f2141f366086a6d29e376bfb86fd0146dd2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/metrics.json": "3b297cdad6dc103584d56442df3a7f2e3331a71bb24d98e13088d2563fde8f7b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/facts.json": "0f6166a5368866541bca5c4ea1754cc043cc7747acb6e0f12df8475bf69d5e2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-17T12-08-36-761Z-61vh/codex-inline-judge.json": "6dd4a06d461ea8b9e32859b9ba99a3a83108845ace725ae1ba5cedee66c7e1a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/result.json": "2c5673aa9e7524f6dc1164eaa0a5e101fa39ace35f8b4dc215e4d1287a7074a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/score.json": "8ad13444985bb6f9ec469e052583936c379797a0166f4d098ecb5d14baebe48e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/metrics.json": "828c520bc1c5ea34460efb68ddcf30e3542b2e1576802f5b68feda7819aa0b12",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/facts.json": "bdd4e61fcf8c716d3d99cb36ed46d3b29f296c69a5381271cc0055406e844466",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-17T12-42-16-316Z-feqk/codex-inline-judge.json": "0efaa1f850d110a07816d654bd3cdf426c513c0085d87e768aa4abfc5b0f5bdf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/result.json": "db9d2f7c9a80e4f979dc7190c8a017012e32f7e246e01b241262ef0ec4077eb5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/score.json": "c64b67df772123e4dae5cece5d0b11d48c866f1b220c9865dbdf2716ceb13714",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/metrics.json": "ead278e368fd2e6b211470bb33749c10837c612a12120f32e3a46748d10d4777",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/facts.json": "f09e33049057684a6c4decd553d708ddc73825b3719b91e05601bdcf072362d6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-17T13-04-25-750Z-qxrw/codex-inline-judge.json": "2d4b566c8637614f29867ec6340e84a778b328817737f8336b4ca4ecab003e65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/result.json": "63ed5e938d84d99630444e7c1dd6c27fd10fb6bc33b69274f0818d62a26f8450",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/score.json": "b476461d3642be7f5f8fb8f529537498ed043e6409293335a7b5364d997811df",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/metrics.json": "4c0456be5bffffb245b1488f981df4821b1610c9ff002adc0e12245ce00b8b38",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/facts.json": "dea4059a6353c217fb2a4e31c877496e500e864844de153abab575284b1884d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r1-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-17T13-59-38-126Z-oa72/codex-inline-judge.json": "1f24c031f04766b28287618b7ce3cd868bf5d3aa06b842349d14a7b7c5b6ad0c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/result.json": "c924792c81c4c4fc0b1e9f1a391a27d5ab76e3f3d84365496d9d5491d8b9c031",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/score.json": "9caf4a1093cce85b6e22d4c20633165131d148f410f6e05d2eef50b4cf3b529f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/metrics.json": "8d92288719f87089612aee754e64bcbbd006d93d693e1164717fcf1b989ca772",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/facts.json": "01719607e811be6b7bf9b95f09e47d0b2e419caa77406ee7dd255e3d8f6d0e53",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T00-21-46-937Z-8qdm/codex-inline-judge.json": "b748fdfcb7351563e2b7484097efa8af8cf60f2a64322a6857baac17a77e22ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/result.json": "48754542175012dc0afa20591c8ffefceca82fb9d55e3b0552877ee03dd3a8fa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/score.json": "0355846df3c3b542db3ea89f3f433fa2ff58981026cd4c25e9d29b73e75245a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/metrics.json": "b365ea7196aba3c866ef4f0d391094ee6014fafb3e83a282eee37851a4fb8623",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/facts.json": "f6483f0fcab0cea90d826fe478706dc32bbce267c148912430d8b17c9a2336e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T00-22-57-301Z-afik/codex-inline-judge.json": "b51df9ced2310bc61b22f6dd04113cbf9d54ff6d57934431cb1d360db232e08a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/result.json": "618a1187e4bce590ef753656f1562741347b0111b1b9dc33175cbdd52a779102",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/score.json": "4e34a3bbf4185c33a2fe325983cca0956675a9fab364e203481176fb0a667aa9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/metrics.json": "56a4062fbb49a25821b9e8891513c8a40052530739c46fe334219b9497574e3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/facts.json": "9276b9bc45241365a316e0ddf38db89813d40bbc6aa66bd89f6be6d40c8e0e64",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T00-24-37-806Z-gvjh/codex-inline-judge.json": "af11615a3b30baf3e83b10419a1c701d7b5b2db2a3e1de99301d3649885cf07f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/result.json": "d3af34987dbbe926d4860ca208903cd7174a693e553310d866a3b05b3621c3bc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/score.json": "4ab3bea921dcd56624abc5b38f7006a7f08cc4ae56c77ea8952a0f39217405f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/metrics.json": "43f16b22377e944ba74116b7b9863dce25ac9f14ac66e2d0dc5b92ef5b42a5e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/facts.json": "745004fc9248589d0631a6d99414b37830939eeb339449f663e0f83f8e053eb0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T00-25-59-072Z-fkf4/codex-inline-judge.json": "3257d49264c98305588077bfe49d79ae8d2171238b103c06a86b81e6dda59ea8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/result.json": "e04a70ae58d1a59b20c95066dcee58179451d1b2a32a2fb07d854724514f660f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/score.json": "5b549c13884fb1456cbd79b9d1880ed3f50872da1af84b7c44faefdace9259fb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/metrics.json": "a27be5933e3e25762bce09e3f9d59babfb799b40a05b11a60104b079680b60c4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/facts.json": "e71df7e78a6468e56d171655629d2090e5563692465bf07f63a5794a91bb21bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T00-27-28-002Z-iboa/codex-inline-judge.json": "55fe2916ea92d40e3e296261e1b1f11f1f014456e48d820cc585129c5814d5d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/result.json": "9259df1b5011c506a2af528db9f3d7e183d1a4ed3d53d79ac939dea2427e64e8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/score.json": "4193b77fff89e8e020a2414138c6fca8f44fa2929ab49acbf7b78153dff96c10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/metrics.json": "768f40373bb6e0bda5163d59b5e1e1f221ab5d654bcf0c8e6e9eb49bfb59a95c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/facts.json": "afcf64592e0c5b0fe5fa734b4bb17f209a5461ccf936d95d150d083e17d33408",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T00-28-00-160Z-8lsn/codex-inline-judge.json": "fe43170f7183cc0e91fe60681f33ec249cbb1690246e9cd51704f8ceb72fbc4c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/result.json": "be07e6d399a74537855981b4d1c2a3c6adfc52c47958cf24f566a0d5fbb9dcb1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/score.json": "67cabbab0acccfe67a3b0bac0131447a387a716bea0767601a019a2e657c2e09",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/metrics.json": "3578e6cd1c54dcddae478325303198f677742fc5124b9cdce9bb39fa8bd7a94e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/facts.json": "6cd3790490fcc6aaa3f23b9aee62d343a9ca34b5cbeca5d5f4da1a1b622f9fd9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T00-28-47-609Z-g494/codex-inline-judge.json": "2ac3793cbf1aaeb8a9fa81fc86109b4d92e591872fe00ea2f4f5b44c14df6a59",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/result.json": "325b4959699bc701e6c3b68ab059203fe6af6e52f7c3836d33d045f90ff164fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/score.json": "2e0e62e002fbd33ba6d2a3ae4e9d17887bc158e27f819aaf1be1bc39ec8ac481",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/metrics.json": "6a72f27e76bd8d452a97c874401a7495b6dca9e5421119ab844beb73e623ca59",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/facts.json": "df5d0ed090e2f6d5b9ca017781739cae4cf55f3994080f8ed89c0d533c54ee42",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T00-29-54-705Z-ocfh/codex-inline-judge.json": "f7f085eef466731c2f007a3f0a0b10b0cb94f6271e4c135c20426e882b285322",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/result.json": "d1c582b7dc74e1c41305ffde3f5c1074cbbae90bc3ace05f5dd63054422da80f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/score.json": "05482004fdba896478ed813ab9803730c7c24b8313b0337103ed3b8e3356291e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/metrics.json": "1510469faf18809f562915c92d4b7de22c0cde29db67147a3587251df6e1747a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/facts.json": "5e1387b630c68c3c4a96b3da7062f6a15a69db80fcb6ee4c2c9f31cd74be1583",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T00-33-42-535Z-faem/codex-inline-judge.json": "f47afa5706ef9906b36253d9e6925f2fdf2c57f7cc5392dec7952c49cc2673a0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/result.json": "986e11bed05ca6e35013fb0cf44cc947c12b476972c7114b1b97fa1f2f4b8530",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/score.json": "d91e7ec1379129d5de5a3c6b5d6eb500fb7d38cbded833518335f408626d28c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/metrics.json": "d8ce4952899c0ca3c237b135a8bd05d28d60a23586c7b72288c673f4beb1fb88",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/facts.json": "c2936c5a6913b0eb64a96ad08df309b4820530a3bea8ec8f1d1d190576c29961",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-34-34-539Z-d825/codex-inline-judge.json": "07f4fe9adbe5469eeef76c99772aca54c314dda9bee95d30c729815c9978e15e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/result.json": "33f2671bf20798a69b5a6c0dd90efd53514ee8b668bdcc81a4e392e286bedfe5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/score.json": "484187e7b8c63cd5f7c5c2557fc0149622265c67e4d6f79f1f86873e98929bb5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/metrics.json": "d7686f7901a027ef177160f01ddc010247e6db14e86304626744f1b268d6a8c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/facts.json": "967ed3b12af8cab350d39b54e6b02244c291820fea25c10fb3cb07f24cc90cce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-35-15-476Z-hk5t/codex-inline-judge.json": "6065dc703c3535c67fd6e73b6de173b74a3c5292bdcba7ab4b3ff9ef04d44879",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/result.json": "73518059cc29722ab832347bbac3fc9fce9a029395f03e7d7b10bc9e4c718a04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/score.json": "09b153f4bc9d1114ee45445a2cdb525e95eb2c37da799b35ac06cbc7d5591782",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/metrics.json": "e2d067abe0b885bc568723c64d86169f966baac823dd4a84eefda3d54d55e386",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/facts.json": "7d605678a0cca681e21976c7f5f36007d1da08852e4a6b9b59eebb5f99fa9095",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T00-40-33-678Z-vujv/codex-inline-judge.json": "bbe00803f73950fb7c894ce42bd9bcd360ea447c19041ac062153b924873e0bf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/result.json": "fa4a83ee85466e8fa6c313328574c1832a8821d434798c20b76b1e5786b05100",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/score.json": "5b077a3e3de25dbcb7da2ae4d268c439577c13dd60e43f800bffcb808989a104",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/metrics.json": "3af15f3491df3ac94bb0f9b560e79abcf9802abd8de59e930fec90a839e2adfa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/facts.json": "f212e43b3702269c1ae05520bdad77e1c3b96245f6ab05c0fe3e13c1d0650672",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T00-41-24-677Z-ecwm/codex-inline-judge.json": "2d03d59f1a0f2feabe2dfdb438dad92ae03e96c0a70c257890f729a3808ae054",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/result.json": "0d30f5ed2d27a889578a3919e8bfa5a325d3ccea7d2e64cf63ce1c34a13f4680",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/score.json": "9721dabfe3f93bef3d148a945a06955e2a32255378a154192ea952c5db4989db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/metrics.json": "54bcf322d5fc1a56446e0a1a2b770e340bc5044e81ac29759b587987b062ac7e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/facts.json": "c1834c8021dac4f2619901f2c0378d1fb2d8087597b8cae68be861205bb2fa96",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T00-42-50-813Z-tdms/codex-inline-judge.json": "b334b40be4abe66f73fe99ea0698c978e4e54d5aab1b8ef47f0426fa4accf9e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/result.json": "bc31d82a08eadc74692927f45d27332a218f996796d18e2c8140fe9d84a177d5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/score.json": "adfb01912b434e72023332a2626997e95dbc15b1b4d89b9be3bf651d3fb4fa64",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/metrics.json": "8442895e07bfc15b5db361f5b83b8d90dbe543a5579468f081f93c715af3a4c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/facts.json": "a4bbb3bab4e32fe3d83c671b94c6c8c396192a7b54837f04c18b520406139d4c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T00-43-31-750Z-cugo/codex-inline-judge.json": "f5191abd94bbf4f647dcaed83b44db921977f7feae8b5e6c1f1dbcf764c21a5c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/result.json": "0e60d119adc286c65d83ec49577a658e6fe4f1e435968e1c95fbacf7b86e909a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/score.json": "08eb8d7d18e3664868d4b043f8a66979214f8830a4ecd000d87d9f6866aa1057",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/metrics.json": "857b93156d2782367326b080b86c38e047179e61221a9c1c979ab052f08c3a52",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/facts.json": "950dd664de69213e855408ce18b6f5cb74de69b137a6aaa7ccbc0009e6cf3a9a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T00-45-39-381Z-gx78/codex-inline-judge.json": "86f6929145096c7ad08558adc7d7c22c88c99eee468b13837ebe1e89556747f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/result.json": "94c2ea786b5410a72b6ae9383a95cbf51d85e61766ad871b7724324e0740303c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/score.json": "351992bea68d8f2d79a3e6390b0cd107ce6f0fe5fb298a061a1f8b2b2049ad02",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/metrics.json": "54a529f37c96291c3d93886ee39a6faa4ec01b58a2d070bc7cf63ad9c8150dc6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/facts.json": "cca1a55a2afd9edadb13445d84a7f23f71e83ab8142a26759149c82e20bede35",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T00-49-13-180Z-y4b3/codex-inline-judge.json": "dae15a9d7faf29573477ba796dc2312af5782f7263c3e8770c76d2ad3482d893",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/result.json": "7468045d85442f70293f7ab890512d2b0ccee4b70576f4c08286a170cf71ff6b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/score.json": "83b015c0223b4032bfdb299248b5c1460e078b1c4a295a2c0682fbebdef3fa48",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/metrics.json": "498ed38a8d003c0438d654ef14ba714c78463df2c086bfa512a38882b444afe6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/facts.json": "e6dfe1120f7a2e3c57e3c7307c75603961d73c40b47dbd881888f1470c7c9b0d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T00-51-49-734Z-lifb/codex-inline-judge.json": "1ebec3137d7831d8c63058b0696922743cb6b770726067288b46cbfb62eaf761",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/result.json": "5fe5e2e043cef9bb0959660c528802f4ba3bf2997780e0d4c4c6710923771ce0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/score.json": "43776e4ce896a606d048f9cb0441f0bbe379fe00882a98ebdd4e5c0f0f5afc42",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/metrics.json": "ba0ce820b8ecd8805db02ab228810587c38e6e79a6823f089ef42d685debbd66",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/facts.json": "25a3c4f4b7c2fdc66a47e6cfabeb650ce447bb0f57a7a9a3bc5655d0cfa68779",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T00-52-31-712Z-7zw6/codex-inline-judge.json": "09f83f42759ecd3b72c2d1021a5e3c6eb6baac4a964e81a455387f148d127428",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/result.json": "f815537e76d684595881ec9ae3d9bfd16d061ba6a5f12a69ae8ec289b10377f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/score.json": "7402bea5fa86a2d365c6bde182d6a504785344e1ce6e3b0fe7b3f8abcde1acff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/metrics.json": "d48724c1b956b63f688d0cd68fffb0349f363ba606f9ce8ede589507713cafcf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/facts.json": "9c83466a3eaaa9bd682c5508aa57dbf455729c996d0bc9543f7cb7757d1ddc96",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T00-55-24-297Z-pn3q/codex-inline-judge.json": "1ba30500c64153fd3246c38521647506ee64795c3daa3f5682e5f84a4127a7e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/result.json": "713fedc51eac1f28a351892d5488619adf5f03d38f0c9ca918165ea0ae664744",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/score.json": "7ae98e6a1c9d919cae166bcc02d0b1190efc9ae2d901fe48834a93af355765cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/metrics.json": "e3cedb44a73a31556dc3baf9f28ba538cb577b8c996be14d5da033fb5630f99c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/facts.json": "3632e8b46b0f23449f47e688fe20b08cb0b3fb21842833b59c3ff1ec8ab0e56d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T01-00-32-551Z-f9gn/codex-inline-judge.json": "2ecff3c648324f2fa7837b68fee3d3b1c9422446378314129aa7afacdd428ccf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/result.json": "f2405e56a7ebf5e3da39f8c77b7c00e958aebbdebd6f34a7c769462980d56379",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/score.json": "b5530eba04f90ea5be92dfecbacc4ec02525fd903727ea4aaabe0e02ea415cf9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/metrics.json": "b5a9118e6abdf4407e2af2e114c0372293f55d9d05e44d4833909c9dfd9d49b3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/facts.json": "36e0c81328e4b14d4bc3432fa6000da6147d0042f697a71150cbbc4175b58bec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T01-52-30-925Z-ounb/codex-inline-judge.json": "43b4805929531eca127310226be48f99fcd8ab85f850e7ce534658babc3ced64",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/result.json": "c3065b72b61997675ed098bbde24e2ace739d53524035b721f7dc3040b9f9bf4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/score.json": "997f53c4c7e0ca3b9f28d94e031c3688f9ed26bcd13295da702265b3a9b43f89",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/metrics.json": "6cfb45e1bd22a4d8552955b4f21fc376c48d637726e8c1869d8642fb6160d9bc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/facts.json": "c66669704548ed4b6d68950d63a87fb6ce06b2f106172aabbde9259a990f0838",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T02-04-41-113Z-jw2e/codex-inline-judge.json": "cb42d8d19996e87d3174e7e689420974f4274f5dcbe73f5fcec5f3d6197b2927",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/result.json": "a835b33cfa3163b2b5b80e8a579a3371c5e7f7cb398e2f347bb44511680326dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/score.json": "a3e4a4093b9c47a9ae029bac6f0f1643d4317ad1283fd14a697b4f186b1eef0e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/metrics.json": "f9684ac82cd62510098400fe7b044b8131c65038f26d978b6f789e843c59ca8b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/facts.json": "2202fdbe044c0e198974c9d52f743989219fa3fd0c161b4a3c12e3167e224070",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T02-07-48-705Z-wak1/codex-inline-judge.json": "9913701139e51e5e6081f58b3d70e1042fdf2e26c3a599dd720c9b1156cb9f57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/result.json": "2da048b5c2fecc2f632d8aabbb78cc18c7860e83a28112200c78b147a416b1c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/score.json": "6e1142b25e21e6ed4c15de92deef2baff677d7d5effbecb31af86affc3eccbe0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/metrics.json": "43a559fd8f2acbae32e770cb6e1bfcd21dbc2d15cb13776cc4c67b9f091f163c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/facts.json": "d074698ac075e2310c72ff7fe07dc9e1ab409bfb85f629c49a0119f7aca32f06",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T02-10-01-055Z-p3qk/codex-inline-judge.json": "f209d71b6f6ce64a181a2fb0af7fd070bd01fe3c7f6e9d94c081a963b4c553f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/result.json": "ed1303260054907dd7a70e7f448cb2c9db17b6948c1e7e53b9e79a63f47d7fbf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/score.json": "902d6280bc66addc78cbd9e157c866d434a2b404a70570edb2504dfb72f6d631",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/metrics.json": "e4e010c1359421317d3ea94fab0c8ec51780c399a7d0233583a9ec19030c8402",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/facts.json": "1a9c1cba4d97f16467ee0f17499cba030601fafbd2c2305a6033b8b99607c4d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T02-10-53-030Z-v2nu/codex-inline-judge.json": "4709f9302020f3341f9677fa7e16aa013df75310deed6a8b95d8aa71c9fff81b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/result.json": "7b303e92f930b44852f8ca3019bd3addd8767733e9a5e48d6e813430cfc7676d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/score.json": "6e32f48f9e7fd3165ce36ef361f404f609a5ce7d8a87d161583a878e66bf6a7b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/metrics.json": "7b0a4fff55136c86b905445fa39365d2831ea1510eedfc791f9cb92f31b902f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/facts.json": "885443d6e86ec7207c6e249374cacef6cf154acd2d88cc88b8b4ce95532766de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T02-11-29-927Z-fmjx/codex-inline-judge.json": "b422a61415dec03ca8813f8320753114786bd07b0afef53b36313a739cddfd90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/result.json": "acb79c8b2c724264fc7cc9c60c35ee13f4371e4641caee3a72b5e728ebd94682",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/score.json": "868b9dac0db312a427d4ed58c2444cb614112fed4139f4e117d2a8b13a426ec9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/metrics.json": "0be4f512a7b2cd3971a0c8a0a8f55d10e44d85756d8f4bc5cbbe48d8bae5fad5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/facts.json": "65d56543acd0812e7a0979823fc619a7b98e6da258dc1dfae4df1c1478ced6b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T02-23-52-644Z-o7ib/codex-inline-judge.json": "b6e44ec16f8279baad1e301295c4db8a22e6e13005a9620086706843de0c05ae",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/result.json": "72fb639beeebab58c50027577df7160871d7567f0a7e1d2431bd038b15a7aee0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/score.json": "9da04abf07dc073b1b0f347e710c783720653563a24acf8c2514691b8d73e61a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/metrics.json": "07dc57e17888fb60828159217712ba703fa8c302bcb94bf457b7417810755822",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/facts.json": "d90af66ef842488f8e8b92c052f6772b103767456ab052478afa327e26c35894",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T02-26-45-139Z-13yf/codex-inline-judge.json": "488bd7d9ad145cab91f610d403372123e2ddd5fd45e187f7876574f796402814",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/result.json": "79210bf83fb2d9c4937a4fe05ea06e876841d55105d3be8c31329193de9d35fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/score.json": "284b62c88861c86906dcea97485f15bb15a68ccb4589c933d3d7e22a7ebe51fa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/metrics.json": "5bd099fc2be023d03ade45e8f6e05220af10d3286f0a41156d061137fd510a54",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/facts.json": "4f37478f268aedc2141e89dea008beaedaab193ed50f6207fa257c72bdf76b93",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T02-34-23-846Z-t7m4/codex-inline-judge.json": "e9d35d36189702c227d75c8d16e2c5a131baf6279b3d1cc1db3c908f093a7352",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/result.json": "2b5f398c5512e70ff373279506f895577e1a4e66409e4f474dfe9f9657e72d3f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/score.json": "cfae16c083d29c2bfd0050e87a68782b7d63c35e9c9ca13cfab35b70bc25580e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/metrics.json": "173d02953256568f5d51d714a33475dc59cdd61a0b6d36ef134208cf2cea33b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/facts.json": "0d147d230f83d9e3ec9ad4eb4ba2677acaec0b1815c7b1d27ef7f11dadec4681",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T02-36-42-310Z-erc9/codex-inline-judge.json": "5e8e19644b87de6135da624a845e4d39cdfaa72fa14b9ec36ccb72333f57d46a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/result.json": "00a0a1aa95a818b898c7aef35f36f7673638e0141f1cf0816170e58180280cce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/score.json": "e8c73e7366482711742a2d31b9610f95f2f4bf845c2e72a8e62acac783614d95",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/metrics.json": "3d6848acffcab6aa13ee6701291cd2b0a7f4c1709aadfabb5f4e0f44b0497713",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/facts.json": "5af4f78dba57ad70d6a0f51cebe2740bc9dd70103b953b28bd42040105a0c02d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T02-43-10-742Z-9kur/codex-inline-judge.json": "3ab447cc76398bf635bd4447a63bb5feedec5f2deb2981680da814d303396ec5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/result.json": "e93874daf2e3e01718bf895ce774e2159b8cc16630e810adffd7190b7b39b1f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/score.json": "21e8cacd48ee0806a30cf641fed7a38a470b84f4f6c103250ed79beff0601643",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/metrics.json": "36198ffd92b95c34ecec65f1cc7669a32fac1268a0a4a59cdd204e7862d619f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/facts.json": "2de3b2763e7086c551a81bb2aa135ccf0eddd020013498e45395026a081c809a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T02-53-30-651Z-378w/codex-inline-judge.json": "8ac8848268c046bcb3be92bc46a6c48e6817b62bc6b236536efdb4277d6a3c5e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/result.json": "2d53fe69ebbbe6162e570d1201274b99602e28ae2d1ebfb88d91d7562908c0f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/score.json": "cb423cabc01144ebd7c1e6b803fe847cf9a332ba13f4eced5cff8b4682fe9e07",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/metrics.json": "ab17239551012ab536a22e6f23842f1800794572699b5dd6d468c68904933c86",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/facts.json": "f09cb65777c21ad77e83681ac5ee3c77d2d91420f39f94b20bb2aabbce4c3874",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T03-23-39-841Z-645u/codex-inline-judge.json": "3c7581c22a8df8018b38727ae681ecdcd1bec7fd9c9225fe072d8a7be8c017f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/result.json": "8f15ca11d247fa3feccd5b1e0bb9388d4e1c61f6054b9677a1ac8c10b3b4c212",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/score.json": "977e8cf07eaf0931b8d49f5dc8f9b469485ecee314f2a643192c1c1c5300d807",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/metrics.json": "f4d7554863e4432277df8a0452d235ac3b80357fbe51052cff2a1177e383fdc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/facts.json": "113a3a7fd3525f45dcec849d330cc5644733a381eaf824d9d492ec0ae5f94e90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T03-29-01-457Z-0vf9/codex-inline-judge.json": "71287049dcd27dfd1065184231f15afe72694ccf7bb13d8d79471273eb57df4d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/result.json": "0d44d27f5263a5f01a3f30126ad5a758a64d84f1f19df13c932100c477a16d83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/score.json": "e3812ee1d5f79dcc2e1a5d475462528ea2506aceacd53a0fbd125bc034bcb4cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/metrics.json": "683e011d6fa4a013d95862c998052106d0fd7d51e161dcaa17f4f49eb05d0d22",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/facts.json": "f046267d3df21280b5344f41a9fecb2b756cafb4d4587a0312b5b3ca0ec24a12",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T03-29-35-370Z-nh1i/codex-inline-judge.json": "0757e466fd65516e8375168345123845efe58b214edbc2ce354ae0e4f9f9a6fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/result.json": "2be13444cbd07c2c007220e0a6633aa85aed6cc848d9f6500e4c5c29fb9131d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/score.json": "cdd743ad73cce7afa2869197dbff70a77cbc56f0ae22358041f52e62f5fd3977",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/metrics.json": "f27506e7295e13313e8098572082eb49c5b8f22c2e20d7b44c8007e2d62b0c29",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/facts.json": "311019d0aea4030f4969f734eec16f1172d16012c5a5d40d4ed16d6cc2a137a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T03-52-57-982Z-0jz9/codex-inline-judge.json": "e10fa542382615f85730c4e67de702246c0cb4fd33c7cb57b968574dac7080c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/result.json": "9ca23e8009504fc164e0badb71419fcda19d969f52514fa4d6ec9396ec0b43df",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/score.json": "673dec12decf1a5c5942a58cb3b346a7c04a55451b88179b7faef21e6d962987",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/metrics.json": "93579fb7e29514a1dccc86eda6a6423794a9ed8eb7d6d9506c17f7d90b3bcd8f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/facts.json": "f1612f361b7af6ccb4b4628d43428cebfff60ce1bad2707c795cfbee34733ec7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T03-57-04-871Z-fp5e/codex-inline-judge.json": "a3aa2eb735187f084826efcf2f6761281690262bc66b3c1f687bbea75dd759cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/result.json": "b488c8d7c017a28b04bf31b02a83ea6d04bcca64f3b12f0c70c5cc1187530588",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/score.json": "4b173f0f80bf954fc6cabf1a0beee356317f2f1a0de42b267fe6a9b119e3245e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/metrics.json": "9ba2bc207c1db7463753c9cdcf229b2d046917bcf8ad8db49c2f6b1fecea7b2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/facts.json": "a5ee2bb14ccd02c13818aa5a1da8b555b4cb8f3c5d21c82b872e863a57307046",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T04-09-51-511Z-57r5/codex-inline-judge.json": "3978366e8dc5b3b795062a15747a05930a77882c0cffcb614a4df8e779ec212c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/result.json": "d8075b5813c2d9e27486f70606d34a1590ac8ef8051c005ddc5e925b6633c913",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/score.json": "7fcda75511ea0272aec23225e5264d17b30a3b81ed8c588b79d835b0d4f7ed24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/metrics.json": "13749068d3dd5f435586b1f745ee3d7aaf1a1357cecb3f5f5483934665dae9c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/facts.json": "fa35da13e0adf4a3d76ff613f073e7a9b7c1e85f96c46af51b9241ba9b3b5ec4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T04-15-56-079Z-778u/codex-inline-judge.json": "a93de8e9bef8143ff6649d140dd4d41b9b2f047d0311605797bd0a58da1483ab",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/result.json": "65542a4a382fc93fb0e277bc663de35ed2597e8fed15aff3088d709b3198a6ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/score.json": "0add76abfbf4fd570a99855583499509c4107fd17ef7fbc67e2691357d08279c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/metrics.json": "71a3026c945eddddc55e73df092913b80df085f0f56aecf8679c9814e6217bf7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/facts.json": "cadc014297cc15984829f482be3c3406cdb3d0d8bf03231bafcdbc64741f50b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T04-41-46-672Z-ivp3/codex-inline-judge.json": "76c06a5c21a531d607d6ca838ab7dd76e62dde0ccdb1ee52498359b99ae8b78d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/result.json": "f07fb7857d0426b43fa4f0ccd60043a1f6f764f14de95a41722a2d6dae48433f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/score.json": "1487578cc8bd9707db2ff1510e2279974520dfe15b8ce1e2f7902fe3ccbd08ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/metrics.json": "ccfaa5e6a1e7580b8982b4dfd60e0f75a7839ef6a30b89a4a6458e8f0bea505c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/facts.json": "6ad800220c3e57336a4e63a76cb6f2929337c74412072c9490380127f1352739",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T05-05-32-672Z-gd6t/codex-inline-judge.json": "e92c5ca7c9b5507b0140ec4c7ba4839f7caa7f73c8ae893972657931695f4e2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/result.json": "c2a9d685463f13c14187732ffab04cd385685797aae80c70d210d5fcb1f75cb1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/score.json": "75550189917c3ea58c017c06fb7019e037ec991a8c5a0252ddb0b89dc0175ace",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/metrics.json": "bd30803b5263919e81f39bde892a2e084f33a2ad60a2a3bb6612c4a0bea29656",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/facts.json": "ec7f91943fb5d059cad69c084c15b5dd59437baacae7c28baaf9ddb72d83fb23",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r2-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T05-37-59-776Z-e6un/codex-inline-judge.json": "f324da5d3e4d80eb853221dd8e4e1dc3e513e4d86fcda8010586934301d6397a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/result.json": "79b210c443ef1e4fd395a9154442cfc4fd8bd786cea680559fd4defce29c6acc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/score.json": "17f8190045f390795134cf276101d4a3221e0ffaa0a91a139fa690558388ea9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/metrics.json": "9c12d0431f51ef554729322f4f8387b1651e40530c2c48a42cfdb7662969b530",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/facts.json": "ecb165ac679c01a5286c05717ff89bacb30e74dd78f775d6686d9194f8b9e8fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-6-35b-a3b-q4-2026-10-18T06-06-32-946Z-3op9/codex-inline-judge.json": "ead08233dfe8bdb845d78748f05897afacc9fbb47a76e974ee3dd2a17d019861",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/result.json": "18bb4d26d50e7278fc4c05a2fe5c19c84f2d6e06689e7a7d1eade6f2be49ba59",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/score.json": "723d42bbe9c6ad8efb4ab8252d9b770c0a67c3098d2251215c039b9428b307db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/metrics.json": "1a256c7452bc5934a851d7c4856368869165011e133810a6e51c6e55e90dbefb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/facts.json": "82e0970642816eef9db94e94e672ce51be35d28b28f3829df2447c22c7a308ca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-petshop/petshop/petshop-qwen3-6-35b-a3b-q4-2026-10-18T06-07-48-207Z-5g7x/codex-inline-judge.json": "a8e2b3bfa49bbf6b9850e26279178fe70a5330c3a517d757c5426dd35eeab43a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/result.json": "275506c0686edf846c3349499baaebf8a4cfbdd0fd91bda777edc750bb7688de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/score.json": "0b21177d99dfd0250ffa234be5801acc2670ea6c3faa4a154ce647d59ca08037",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/metrics.json": "96b6a7a7cfd750876d73eeec1a68946e394f8b52b49effcaa40c9e33e456470e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/facts.json": "5ee4e7ef457f03206313cb372dba3c655af25dbb9f59a472c1ab0fa7bf5ba4b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-6-35b-a3b-q4-2026-10-18T06-09-28-480Z-lwhd/codex-inline-judge.json": "e9055cce995597894dddad08084e2c8c403a0e2da7fd5750277b43c0ac7e469b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/result.json": "dcb330f58a8330c7d7990437193c518d8086c4fa5ba901974f28ad72f91ef316",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/score.json": "cae7602ff3b6e20b6ae7aff18c7606c543216ec4bdf0bfe28f298412d2d73470",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/metrics.json": "f5329770337188177cacf83cb7d27bfbb03f04e0e49246b584d95af65325a7cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/facts.json": "39aefc5a846ce77a9b9c434326b396648ecd08218af87f3ff53023da135bd470",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-6-35b-a3b-q4-2026-10-18T06-10-59-651Z-oe56/codex-inline-judge.json": "4a2bc86fda96bd2ee74afbb4201cf51ac7d9dad3d53aa59d2b3db0c86819942a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/result.json": "f30e705286afddf18ed2718dc70055a607643999b45abaf74c4aaf8383f9494c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/score.json": "3676a89e5286daa085f6c5eda9513b4b2b30c74d1e511e650ecd85e918403214",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/metrics.json": "c75000b2391872eeff980305a17b145e29d5857c93bfe17a9c743b5a127192dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/facts.json": "94282f89b872d260c38b99be78c4e70760e5c1b4c095438624e4a26dce8f34fe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-6-35b-a3b-q4-2026-10-18T06-12-32-028Z-pnna/codex-inline-judge.json": "ce6535c530b5ceecc9c5747e202faf766dec02174c5f7e83f4c957acc34fc5b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/result.json": "c19240f38595f135e356929ddead80ddb71099bd4592aa481cef8b682293eff0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/score.json": "43d940f24d9bf8c0ace27b51eb673a0f45f2a2ab5898c32245e5fef262667b5a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/metrics.json": "ca5680e95edb9c0861b4cd41d8f06ced44435416d8b9b6b33052e6554c9cb54e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/facts.json": "f331baeeb81129aaee391ec106a069ba32cb7d49c810df625620a85c2b62de47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-6-35b-a3b-q4-2026-10-18T06-13-04-084Z-iqgt/codex-inline-judge.json": "844fb25d513a4cc9d1f58569f8dfb5177119b177fd7cf6ce6a5a1260f3a40493",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/result.json": "b66eb1f82bf0adcfc0a359d26848afcc6ba760515ce08398a483a16f3da648e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/score.json": "0342bf01eaebed533bdc16390b442536de8e407627a13681db57d5b35f213a3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/metrics.json": "b2380f35458e6caccbfffbd8b964994c8d28e5fb2fd1f88a1ff593be02a5b66c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/facts.json": "0cfdbea8b70944440d031f2a20ffdd0c7e5c94233bfe9e7d431cd59f7e5832c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-6-35b-a3b-q4-2026-10-18T06-13-40-290Z-9o5r/codex-inline-judge.json": "92d8110fcd114fe77510436685883cd005c60909fa01fe76afc28790f0f9ce72",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/result.json": "266eaa132a83b162157a1b507fbabf3ffffe82be6c0dcf0dc1b9d76827247154",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/score.json": "bdac12afcf5627ecfb882432f5ba105946ae4f0b77df2b243b405e35244f9217",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/metrics.json": "8018e4ecdf27e33bef928e8c7652553be275a261488c511adcf23d15c32d7f51",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/facts.json": "0e8654a8c2f10226f7cbd90e15c89182b5b334a00d0561f2c55f8847c3ea9aee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-6-35b-a3b-q4-2026-10-18T06-15-52-609Z-apu7/codex-inline-judge.json": "28c6999dd35dcc5de5d4dfe441e49c2b28da282a213da4562b77762cc469e642",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/result.json": "e81e924f0470cbe58d7d39a883901fb26c22e7f3f27e7c5e5749d8797482e275",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/score.json": "d4687a1d9370e67a410a3ec74e5db8c0acd4de60edcd2d6d04c1632b0a25b879",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/metrics.json": "cb63296c2e6af764f2774160ecc48182debed1ac42aa58cd14e701db99b5c1c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/facts.json": "2fd63be2bdf2031c46f7fa8b53cac274473488d57fd6c00b7247965fcb71c997",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-6-35b-a3b-q4-2026-10-18T06-19-25-295Z-dyus/codex-inline-judge.json": "4414ad6e3afa4a9ed0b6c24dbebe710c7bf03b70e90a638c8d54c9a7afb9e07f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/result.json": "78db5249944b5cd385275100506c15c363dc3a78abbab172b21397edcbd352a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/score.json": "7a27e3e415564bc1c2f639528759bb8492c69d69d8391574cf18ed8ab391e4b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/metrics.json": "b1bb3d1fa4a6bf8728e192d5514b24af9301f6e73b802dc2f73e8595a2541b8c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/facts.json": "7703fd08a4f3f7b62e2f079aa4bee28b7929458f32b34d60f703ae93988952a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-20-31-358Z-wa7l/codex-inline-judge.json": "f54c5601ac56cd759d0f4913cf73f4a0cf6481945f25cc2819f4658d4f235239",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/result.json": "e4686af3de3119b328ed6c56cb9d795c9fa62906f2af8b97966165af2e8dfcd4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/score.json": "5ef7913bc6fd0685ae83e5a120dbae95838b4da0e8d2ae4dcee34781541cb7d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/metrics.json": "971db00f6719689ac13bf7aebb7999ccf35a2b795a1379cce7f4fd999a599afd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/facts.json": "f0c9832ca2ea03d8f76ff5430d7ba36b36d198a2bdaab73cb1243f9913f3f848",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-21-12-316Z-3lpl/codex-inline-judge.json": "0b2efa86b7dfaa7df4e2b2cbb64d086e74caf67a437e31a174b4de39bd8a2719",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/result.json": "917a2dca08ee57f73591eecc623da0953d97c462321eb0dd955a8d66933ce065",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/score.json": "e7d348a1b1657ce4c9944bd044fc49e46000d96a40b6bbaca740fd35c11972dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/metrics.json": "711118c68a7fda073aa93d170e890c5867f1715339d7359dd2939d35464ceef1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/facts.json": "604f63be33a6839dfcc76eb78cba82b31b14291e1e38e8656b627bb1e4427e0f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-6-35b-a3b-q4-2026-10-18T06-24-04-886Z-0mx7/codex-inline-judge.json": "a52870e2551ad5b8aef45b038589d3aad3305326941a25ca1ddf458817ad2660",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/result.json": "c60f6a751ffe9780fdda5ae559ce7bd368fc30c8167fd91a71f8979c1e1852bc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/score.json": "ad174d2404f2481ccb9243fbb50c00c1a59a0cafdb2c8ec589eb0f2ee1ecafed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/metrics.json": "87637c16d68bed73195cfadc987f1890910d93f8200a69386c55611ba8bddad4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/facts.json": "bde6bbb18a1bdb7241390d339dca60e34c06a2594d5d84a18dcc0e09578d4f04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-qwen3-6-35b-a3b-q4-2026-10-18T06-25-05-949Z-4t90/codex-inline-judge.json": "ac92c1c41673285a27d8cb7ccc1ea9e8bef6e283260cdaf04d9bacfebe6921c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/result.json": "334d847dbc7e8eab955a6f46da1858dd7c8e9d8e6b05f97e6970c771f305c967",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/score.json": "2a2959029b945c5e50e79ece09c8400052fc77359ee053985415c3129bb73831",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/metrics.json": "a2a1b73d24cd72ebc465553a18e4a7d3b20338ae958fc5f849ed0ae1ea360a32",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/facts.json": "500edf627faa8266344101b56106115db780cd421ed15e95aff14853a866e508",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-6-35b-a3b-q4-2026-10-18T06-27-17-238Z-qrsi/codex-inline-judge.json": "b6fedfe621880423fa6d23fd7007a9f3fe1610faf53be08dd5ecc33c58eb6ac1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/result.json": "eca20cc9e93b90a514433d0038494dff03960424b26f3d54b7b99ce8ba98cc01",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/score.json": "561ee3079ad4ff69c9ca0e90d6107b2f95b0b7462d71a8a808a5a827cff411ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/metrics.json": "3fde24e125232d0be05e452a9bb5951319298dc818665c9b9bf3126f4da3dcf9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/facts.json": "5efa0715141f154b97c7fe964206eb04054c0872db9bb459f04561aa132b4bcf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-6-35b-a3b-q4-2026-10-18T06-28-03-195Z-cd5b/codex-inline-judge.json": "b4231ec9e7feb89fea968b5e11815025bf07f9f8f1f9d4824bbb5c9c150e32c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/result.json": "4651d9c786c0fe36172a9a3561df8fa638bff21bcdcc7a85b49d3f8cb41e9347",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/score.json": "26942215625398ed2eaa0f273fb4d6a2f63e9515b55c96d969f75b1b45d5e6aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/metrics.json": "9959ed90a561b63d08eec523871d2bbac0dd83243a44471fd09fa2fca1fce3a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/facts.json": "776e769bad86b6dd72ce8cb65f70050ec69c91423510ea803e11c29d56597ffb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-6-35b-a3b-q4-2026-10-18T06-30-04-747Z-u7to/codex-inline-judge.json": "306afbb6b7c3b08de6271dcb0934892f6bcb924b65b1bc87808c6dec707b2ac0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/result.json": "201e0c0de33d6d83fd82b84e9184a3d62aedd825326b0c0ecb46468c6276791c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/score.json": "597b7251d8e1dae267b6c0bf41da5fe502cbb15627c7a9222d1c316cef4cc7be",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/metrics.json": "69a3e9ea10c44f442fb5cef638ebfd47a08aad5e826398e35f378b1ab89dcac6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/facts.json": "8be8b6be54208d39aa02b6e81bca7a2028447708448aad0305e086a6a22cd4a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-6-35b-a3b-q4-2026-10-18T06-33-04-813Z-zbo4/codex-inline-judge.json": "6d449cc20c6290e50e16b3f484274dec38dc0227218a65466c325d08637bcfd8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/result.json": "24482b9422ae04868d5c7c90bc8991689c04b8759d325d6d77097f3eca2063d8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/score.json": "b80f145f77de0bd0eb566e58a8f0aa3ed21c1be5ae92b1a59660ee2d2c1a3978",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/metrics.json": "8f5e6e71c9cbcab8a16be166f84cf83094cf52f250e0c5f6bb533797081553d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/facts.json": "bb41e282db211e8c461d452b3440de555be334e8e6f10cadf98bc8242b476bed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-6-35b-a3b-q4-2026-10-18T06-36-52-778Z-3jrm/codex-inline-judge.json": "303fdc42035b12cf2d99150c23d008e3b33e1e3f4d40ae9dc60ff14545299dc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/result.json": "81bedcbaaca970625bb97ad65340c3d9e03c1f772e17451dd36a6b73bd8206d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/score.json": "c269a54a1927f94265e634f7ad890125d8a5d14368d15c4e38521854fe8058b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/metrics.json": "b8f8a788218be6182686396b4b484e20420f4f774dd1ccbc4c23dc2e26fde432",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/facts.json": "1e18c8356b111498f6a1d1f54f7de235b80ac587630d7f760051df07a2d6a475",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-6-35b-a3b-q4-2026-10-18T06-37-38-769Z-pj7o/codex-inline-judge.json": "c594d178b42fae24307072189e985477815cdb564c0c6ce3de6ba4a39a6ef0b3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/result.json": "7893de0136063d4629cb5916b64d3840144702c28b715f45d5e46ea1198e078a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/score.json": "041eccb066f76f91039e89848b23d5f5ba2876f384b0013201a7933320feef0b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/metrics.json": "49c37e4505b41ec370805fc49c5d92b542e4f39bac0ca79e2172a07800cc0feb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/facts.json": "347f5a7e30f09075d6e62b62653c4b09397d51a6a29897769cf2a51fc4ca903f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-6-35b-a3b-q4-2026-10-18T06-40-01-280Z-xdgh/codex-inline-judge.json": "02764d9c0e362e2c2b7ba491ae53c7f9686b0676ad80adc3257755ff5d592561",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/result.json": "c40541440e581d07f94790945739c16b5b0db805214fe3cc18e2db4a560adf26",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/score.json": "bf7297a4ad1b714f652a968137483f8699d5403045cf3c9902cec3210ce073de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/metrics.json": "1f8706500843d76458e696fe772d9732467485c23c4e81c9b2e1b01c0ef26f6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/facts.json": "c63aac2937c20734db759a3953a04723b50652b277d1b205ee8efd31853007bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-6-35b-a3b-q4-2026-10-18T06-45-29-598Z-adxw/codex-inline-judge.json": "8a927c59afba3ef13625cafcb315ec0dccc43b451b34bdd94d1f88ee8be3da00",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/result.json": "23a653c5d51eb99e1c946f32f96bec81d9de1f361b658fe1e953c240a051b12e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/score.json": "8d502e7d38d61f76a78728bb7ad03161a24747e8c3191d25ca9462f1441446cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/metrics.json": "9fb7f4f993c4f166ff0041a3071787e681e7e9efa43990834f956e5b9269d214",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/facts.json": "efc68ff69af6090d07718359d5d7894c3ede60c525f9825b159262d23cd8b2cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-6-35b-a3b-q4-2026-10-18T07-35-25-168Z-wcir/codex-inline-judge.json": "4f8fcd683f957643c7e58cf40e503dfd8c92783799f3aa93855f97fb40e79ff5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/result.json": "877907235a6cf46d08df157bf9081f8152930fcd826b1b61f051488594d8d9da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/score.json": "ab4e28ba36d29beccd99d04599c008658b18292c209c208f2f1d3a113bcfcbff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/metrics.json": "2574bb74b50e2e393b26d39746c625899c65006f91357a7c4f7eb325ed94223f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/facts.json": "db807f30fbe7d9e0231003eb714fe01661455508ad6df30d2a781ad04d78d73e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-6-35b-a3b-q4-2026-10-18T07-40-49-533Z-zhcu/codex-inline-judge.json": "216490ec047d6f428be3133f882c328612cc5a0b355e10a1ba4367c38389c4ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/result.json": "7eef1e77ceccf8a053ba836d8babbe87327320b5df3adf90eddedeba94ffe06c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/score.json": "1a4eda287262d38bc0d2122a54b9cdc3e6f31af1b4c09a52b1bfaba0a514059a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/metrics.json": "433cca82a27b251165f1ab82200be16c75062ef8868e7ec3d70e16e7bbb4d9b4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/facts.json": "59bb5baa7204a54d9cf6f716a21a3821efa6695cc91b62c6dbc792e873a49608",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-6-35b-a3b-q4-2026-10-18T07-43-42-071Z-3jc0/codex-inline-judge.json": "ba1b3b5605bf48e9f92bc0c87eca3272a552241b44d8c8ac8fa748dea8833612",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/result.json": "0d14e3045ad56f7c2b284ceabb0817250223e7adbb56ff0c08f4324d8e0cb3f9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/score.json": "c04b0bb623731dc9dafa0d1fbc1eba4a22a63ee322a7cdbc95afd40f2b607eaf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/metrics.json": "3fd3656a71206980e4e41a4e7035d79ce74d28e104228d89b6dded0964723a3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/facts.json": "7c013d133962aa8e239b6aafd239b250ee799695c8b5f0bee263e0a533e770e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-6-35b-a3b-q4-2026-10-18T07-46-14-455Z-6dzt/codex-inline-judge.json": "bb4a6996b5e009e83e6065f8b8f5563dfbc0136bc45b697dacf65f755cd647e8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/result.json": "a4445ad77b48f7d2a7ba3a4eb2e228e767c36d97ccefc0a0d9d6aeaab28044cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/score.json": "efa9df24d0deaf4f1e84488eabc196c5232fe7d51c676b8712ad53c530ac5fec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/metrics.json": "80fa3feb41bbbdf690368eba1312ee033043a68b40d9e1f1c253868cdac50dc1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/facts.json": "d629d9f523c4e2c436b66d07041ba9033fc29788d59277c97343ce823a31234d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-6-35b-a3b-q4-2026-10-18T07-46-56-383Z-jdob/codex-inline-judge.json": "78ec065509e58a6a9627c6a7c72368f0f571afd84d31f5aa65fd91292c336372",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/result.json": "b6f7d38746d167f1c2f26fdc129d6fbf0c213df7d89d37dbd1a10b54c0eccaf3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/score.json": "cf0c29a2355b4ff53573b105ba314273696790297262120cc713bff75f47280c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/metrics.json": "bdcf13fe0c8a0739c9970705bff94765a812486898d971b2a3a4e43dda76ec5a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/facts.json": "d287d7f7b424a172ec78699af62a34f9a88735cb4c72d488dfb4077af5095f7d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-6-35b-a3b-q4-2026-10-18T07-47-33-322Z-o01z/codex-inline-judge.json": "43719e33bd580f99f0e9ecc0d92c3543ad7dd4aada50327d154bd45fe619ff71",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/result.json": "f1a17b4987b363abf8d7bb61200609f1bffc6263535bc70d6c4c9b902af2dd0c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/score.json": "d6e246f25db59b5052c169edb77ebfc29676e67fc2d0847c9218018f2978f886",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/metrics.json": "31bf7c5bb4ae2b4eb6d3e85601cf4ed2e6f0619f86de80330729356353d7940e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/facts.json": "2afb7b5af521ef76160cca63a72efd47b673c226bcc75efaf94f111eb69c590a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-6-35b-a3b-q4-2026-10-18T07-48-24-276Z-x59u/codex-inline-judge.json": "aaf5ef16d0d6cd1bfe93f8935cfd4a8d24dc0850900634ab82695d262020a3bc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/result.json": "0de2998e5cd0c6b6bb0c9cf9c233c5ad2044957e56bbd6f434080066ae13c7d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/score.json": "9e5d1364b8a5c69be5febb11d2c824c8d9b25acecea424e3482ec455cc88ad9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/metrics.json": "54f0f4ce4951658daed02e3c6d68d4e01e78a2fa8464011f23eb4add34dcdc63",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/facts.json": "1548fde10578719fd949ebf28c2070b29e41a567fbbcd5fa17a2ee8c3b5b1a95",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-6-35b-a3b-q4-2026-10-18T07-50-27-578Z-ha22/codex-inline-judge.json": "e25b43dd3642294ee24a62f6e15d8db954014664b477d85387856192da0e612f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/result.json": "a0b499e42147f08b9fe3cad23fe6d122f405059f7c1daf5cd0843703c9588cbe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/score.json": "ba7c8113f0937f3522c52c21a1e391a543d1aa1a822b9aeb7adea549f1060ab9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/metrics.json": "4243cb6b642ec29294646788a8df71fbd9f470b3d1d5aa5208a343065dd840e9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/facts.json": "6851f1b872349f5685ce8f8c25b6759bdbcd6d7e999265530022f5169f3b38b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-6-35b-a3b-q4-2026-10-18T07-57-26-125Z-36g1/codex-inline-judge.json": "dd663789549d63034eb8455664b3a1c948a5ec783b387ad6c2bd922319d21956",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/result.json": "930b76061ad4e580a610017140be15f7f35c5e1a340ed933766b149be39bcd0a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/score.json": "37d17e0406ea2e4eb274527e19bb7e4d570cf8211fbcdb4a0e57294b06ff777f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/metrics.json": "a49af3c69572f6d3b2aad0ccf9f1dedfa8687d14cdc8abb7d68fab6db0cc24dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/facts.json": "1dca7254a23a928e225938645ae54c23f1c9820f0b2d88967af98270707d3d28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-6-35b-a3b-q4-2026-10-18T07-59-13-468Z-p3yp/codex-inline-judge.json": "2e52a4ef73e552ec0503a68d2526abbcccd18fb9da033f8b671354cb1598d9a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/result.json": "3004e8b207eb9a022aea880bc61496b171b2e133f4813cc02ef41fce3922abd2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/score.json": "ffef6185680c496d6fa8fbdd83f4fec52a01e9e24d552ba2a39e97f06da2bea1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/metrics.json": "fd9b0dec8fe03833a1a4df0e211560d7bb360e64b23fe71b0093dd7a8d7e063e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/facts.json": "9047a5e24153b437de9cee6fe9406ce0fdaca6abda1eab33d4f1b972a0ce7c4a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-6-35b-a3b-q4-2026-10-18T08-13-25-644Z-my10/codex-inline-judge.json": "4eb8394cc8ebc9f4ff9686b26ae1295ba4d883e1a65d90336836e65d7fa63125",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/result.json": "34bcbf89c081e6009b1109d7b1dd11735649f74946484c1506c9579143906316",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/score.json": "ac669ad99b2c4d329d37190137aa59def70bc98d8ba5fdf51b2ebbad38628f0b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/metrics.json": "2b3fc783390a7a2751a67a48fdfade41f8c1695d1bdef26c6d49a07fe5ffbed2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/facts.json": "d9158fd696e453712fb11509a94b1963e029a0c214c40cc87ce005757edc54b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-6-35b-a3b-q4-2026-10-18T08-18-04-628Z-1eh0/codex-inline-judge.json": "f07c1696187aaa2df992c044356a133ea85db69df1361a0d4851bade41445a9d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/result.json": "fedbf4d3f5da4b651b831cbf3a24be9ee22cee8d11381c186815d417646ed8d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/score.json": "27a376669ed74c8280d608b30f9cbb24aca84252074c2f6407109462eb2119e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/metrics.json": "a40a75813782550f6099ab7e24d8d446d1d0aa321d26d3f76538b3cf590230a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/facts.json": "9caca30dedd0851a00140f11e7fc8156a31bacb3cc389865d97bd0898502789e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-6-35b-a3b-q4-2026-10-18T08-43-21-325Z-6gnf/codex-inline-judge.json": "8e8d1f1a47bfeeec9ea5de9fb457b76118f0e15a65b9a2ef492768dd4e4ab836",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/result.json": "a21fac70d8cbba335da14545b5a45e340320264405cd5506c9706889032258cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/score.json": "09709ee92b75a49bedcccf28b061b1a42ab6a06153e9d453a79600c24ad74324",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/metrics.json": "152b46c57b32c0d56b3425012c8fc6740deafd32897aea60f7567543a8777693",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/facts.json": "da8473cd3120fa86e37985d462eb78fbd16c978b26328b956c0d02261b03e7dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-6-35b-a3b-q4-2026-10-18T08-47-05-029Z-x2oo/codex-inline-judge.json": "df4bed9fa28ee9f006de0c89aa58950e4661abe3b240014c37dd4dd1f0d63ec9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/result.json": "3d80364e9a8b82605130e7b9da2c8ed4559a6d01c53b8ea49af94b9b141379e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/score.json": "7511dc42a18af3ce52b3ee377dcfa6e4ca9c79bece3c080c24415073e1e7f9be",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/metrics.json": "839d6f7f7eb89024c02030e77fac4f28c99b50dd46dbd6f15c60137c44ea3711",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/facts.json": "7c644fd5331c8ee098d101821045acda14ff5327da44c6a331445d1d78e5aa07",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-6-35b-a3b-q4-2026-10-18T08-47-46-965Z-jf36/codex-inline-judge.json": "e4f79de46da2a92cfbc62e7186cdde6dd11c8856f0935be83926495f94235d81",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/result.json": "9a9a17f2a36b02478ae2853336a69e47a4fef60d7e2b2b7461e93bfa8d521458",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/score.json": "48840cf44ed2fad2b3125c71af7ec902260f30a4539b8a3895c9f157a78aef7c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/metrics.json": "5d34188add8a118b5a77c7c63dab1f27fb6a21d26f93f6911b1cd9e4afe0744a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/facts.json": "7ad3137abf3f09920236000a3bdb32c3a5bf52786273532f022979d08563c15a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-6-35b-a3b-q4-2026-10-18T08-53-20-245Z-90n3/codex-inline-judge.json": "ca4d0b689bfc8c48765d7dfffc91cb2b3070ca55d430a29442adeab7cb9d8b3c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/result.json": "004c6e13bb956667e6b347563063b3bb9662b98332768ad7ba4e1c924f8d544f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/score.json": "82f18a93b24ec4c4d2e371c70b5b2694588cc39295b6be8160971c94e52d05de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/metrics.json": "7f55911ac80af536369e09df91e6e491882f6bc576406f56273126895c38d6c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/facts.json": "5e5b12f6479b2c0dc55774bbd8c94bff935fff5ddb4e88a0a81f21fa051ed8eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-6-35b-a3b-q4-2026-10-18T08-57-52-158Z-f2bl/codex-inline-judge.json": "982718ff47d92f5327e18bdc02301478a965db6b75c0ff6a1143b8045c079be6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/result.json": "4aeb53cc86122afa9b787caa5b802c628493ada70079de29eebf940504221409",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/score.json": "b651a1fcf9494a9f35b4a10c6951e018b18b419bf10815366e8aa4ad042798e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/metrics.json": "aa4a3b7acec120dbc17080df1522e5fa4378e6545346cd177bdec9240b03211e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/facts.json": "624dccdafeba8fdb53faaef1c77202ef1f6771ad905b02b8e22fff8aa262cc7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-6-35b-a3b-q4-2026-10-18T09-21-37-990Z-fd17/codex-inline-judge.json": "42e13df91c7ff42e85f77a6ef3c8fccfcbe38f72b220820e03eea5fe30c89830",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/result.json": "c26fe5d825439077dad0366b05c379fdf4f237973d4e8b91b4683d0f8ea9b061",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/score.json": "edde7565100dbdfbf1827aa0df204c93140baef791438180cc898755f579be3c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/metrics.json": "85642766e75c22f6e60c70abc84787fd4dae338124ecb641ca001ee99c912cc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/facts.json": "8349c41dc1b9a3856d85487f68c1f086a6ad696d47858c7aa17782928ea9ffd7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-6-35b-a3b-q4-2026-10-18T09-28-37-562Z-cklp/codex-inline-judge.json": "379c22de5e53411f4675a5f1b61779d5dd1ba11f7a1623e56f6cd16593efaf2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/result.json": "1240f6b3546eb44b0dd48a557c4b7c7ccb058adc76bbbc09410074cb62ac9e56",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/score.json": "98d865238405e5a00cd294049ec3ef2ee3cda248115d42a15030cae506c4d34d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/metrics.json": "c5e118a5451884f5f625e9a061442f1f5225e827c51b6b77bdde24933aa9556a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/facts.json": "5c104da955ed1198575844dcb9bd050596da3b4d2d0ddd209603f18a0c53f885",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-6-35b-a3b-q4-2026-10-18T09-42-15-848Z-cnbf/codex-inline-judge.json": "bae861cc0cd7ce741ac46590b98b4848f54f86fe9d298b6691785978bdf87086",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/result.json": "ddcf3132a73cc1baecb4b11a309079b62eb84903ebc6ab3bc964c0a04ab48194",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/score.json": "922ac32c0e5ef298b53de7b400b970dd5157e2daf2ba867437287a522beb1fa8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/metrics.json": "e8e93d8012fdef67970c4472be4261eaeda388f66e1211d86eee62b17a38b7b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/facts.json": "66299051e6dc6ae4a0f63d40730cff8d66a3b7e4e8b4f1a97d1103b25f775cc1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-6-35b-a3b-q4-2026-10-18T10-09-21-591Z-nxc9/codex-inline-judge.json": "dfbb7b4f1b1fda751c8b5e7abdd4275792b5577a1d05bd0298b61436fb14899b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/result.json": "13250b1a9958180bc3db36c2a56373492a46a4b5b54f7bb0bbfb83ab28f886dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/score.json": "ee05b0a4f3c7636394ac9917d855ca7063fcea9c329f74b9b2b5c9e96c82208e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/metrics.json": "5806eddd50bc9121f7a423fe52b931641e59a480d0c448bb4c0e21b2b647a9e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/facts.json": "47b65d1435a97cf46fe74ce026d05c51d1e5fe40e1e9b3ae4b51e296daaf3859",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.6-35b-a3b-q4/r3-qwen3.6-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-6-35b-a3b-q4-2026-10-18T10-28-50-605Z-b7b4/codex-inline-judge.json": "63cd70cb11bd204fc3a2e52cc43b5b901e59a865e7b3192860381a7295dadb3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/result.json": "c4a4a47d2b1efcc99dfa3475a0190f7e56aae9f152bcbb56afa39739498dbc2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/score.json": "d1d8517a8d6c329421e601f3553ec4ca15bf790ca153d1ecb84a1e1d8bf155e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/metrics.json": "7e1c3b43fc11992a5f8d83ffc33845c2dc7ac7a6542b42187d807f83c63d63f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/facts.json": "9f5b15f6b8b49c5da04e1baa23cb5a5821869994b120562af57bbfdc85e27b73",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T11-04-57-117Z-uygs/codex-inline-judge.json": "1b562d719863894a26f7bc8e52aba8a3a3c9f514e3c7032fae624909b5627f47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/result.json": "1f921d7eb52a5e02cf52e82ec4f769431a3fec6ef68186f8a7af4cb3ac0cc09d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/score.json": "c07fcf9ba12a68ac37a74234043949562490aca71489657992efc651e72f1a06",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/metrics.json": "a1067d045e7562ec49c5a92d03cfc85085787285de11bc3e04e999e55061c634",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/facts.json": "7e3b5db8ef11b47aeb4733424e196f2e6ca9d6df4efc0f5dca1deb319d1f65ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T11-07-32-717Z-5qyl/codex-inline-judge.json": "a87e17c9a31c02c95c4478936179704b0fb111157789215b80d1afc75571cc42",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/result.json": "4b8443be6df64a0b9be7211a2cb6fde52c0b564f5f7aba34602dc2d6f31f77a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/score.json": "ac9091121115166edbe1a521d3d746d9a8bae2657c60de25e92a9cc66a78c867",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/metrics.json": "87af11743a670bd9d4b46986a105aa868781cd3679066fa9dbf4178800e48cdd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/facts.json": "b5d78669baaa426001d4f4668d837fd59c25ce799323095dc5ebd66111d4c89a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T11-12-00-664Z-bq1a/codex-inline-judge.json": "79357ca465a84d4663e4bd51f682cf3a0595a0b09aba4073a690771d4b396442",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/result.json": "4ecc65e5ed5f5383001ae4fc794160f7f8dcb2fcf75cf3eca481c038d4c21c53",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/score.json": "53e3b8d455b1c12eb37057ca834539aefcad64404e25eab38d8866be5792182e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/metrics.json": "cc4c3a1f7f2fe2ed0ac45055e37b120a6f0f7fe05cc9ed9de2d42825fd376e56",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/facts.json": "36e4c744ba8da6cc5e055b9c974457c693e5e88a1c9b3b00a0933a81f206c218",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T11-21-23-872Z-j68z/codex-inline-judge.json": "161a50e066b854bf7fede6a6b3c855a2bfcc29890295cd5e706a0ce72a8f03da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/result.json": "5359ae528799f37aa1630140b5f0deed866c036efd93ffbd6079fc6d02cd8a96",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/score.json": "fe6c12824094bda7755b6eb607f70c2c8e5162ffdf529d6e253d2c8ca0e6e6fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/metrics.json": "3b10a2074aa35b1557ac2d5af090679210b562668422759d889523e8221b2a56",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/facts.json": "a12f0c07f0c60eb77040f51c8fb837aae1129672a3ac041ecc6f05dc17e5b6c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T11-25-56-834Z-pnwj/codex-inline-judge.json": "b2b7d750adf7b2f7c914d40054202b3ea8eb85078b8be401c7695a22c3a3b98d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/result.json": "4e59cd0786e3b95bc7ba1b44c1751fbba690bb0c3240ff9f7a37db6d9d2f6e30",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/score.json": "20641afbf9e8fa01df842bfa7b2bf6bd831c35eb989bf1e7bfc2ea073e208d52",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/metrics.json": "6701a59c7802cfdbd510cf9d9ed06894cb19d3cb578ad98fdc6b02742d13acef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/facts.json": "a36981a188534a4f0c909f2c438f4fafd71da9d0810bc7c5a4c6bc5a629926de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T11-27-03-843Z-zr50/codex-inline-judge.json": "4c4bc2ee24c5eb29d8d53096b09a0689982ceb12353e07d0d97e81e9f4e51941",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/result.json": "7890ac471e2f37d894f6622252b2a23a809154048259b271dd624c0d736f6b48",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/score.json": "d429b5a2ebc4c1b3a7fb9bc6ca026141c9f88bf852d48aaa7402ad35fa28e0ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/metrics.json": "970ceb10c001d504fd6165faa533c62dd27c226e39c8aff9dd51e4d609fedbbd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/facts.json": "3c4a327b72dd7215f15b8ab7774c6140f9c0184166037bae5f063032e236c19d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T11-28-25-943Z-pwr0/codex-inline-judge.json": "ef08de8a6bda803893df46cb8d025efa866f959e0f63dfe693549dd3fde3d1a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/result.json": "01d739ac4f07ec326c23be009be43142ab72e53b444a139ca20434ca842cfbcf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/score.json": "7bb858e8d500bec033e3adea805eb7289dc0d407e69cff3463e8d9da341f3723",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/metrics.json": "d6cca3c044fccbac1944363d6ac8ae155e1aa022980c556a122b4ad0d65c7d4a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/facts.json": "7295fcb545dae4b0c5674d933ee8a198b748e62ca57d4043bc486832557563d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T11-30-03-109Z-8wpo/codex-inline-judge.json": "1bdef2c6bdd88574c622f7a8b717648c6b7a96543be863612e305834515dc7d3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/result.json": "8a8faa4325738aa374bac7494cf7435f662a54ab0ae1218f3e42ef14f2fba728",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/score.json": "ff519e1bbd8f053cd411ea791b873aba5273bb93de659e5e3258b1a723d91581",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/metrics.json": "e5750e90e80227e532d642c82ed8e4cba03d9da1fc1dc4d4c613e9f4d32fb92d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/facts.json": "3f505a885304f4fdf9c2060d4a3314f08637042db49e01e0d008f7751ee31041",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T11-51-55-486Z-mn9e/codex-inline-judge.json": "2e4f04e9d401f12617a63a4c7275f55bf7d649bb15d0a097734736430d461038",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/result.json": "7fd54bebd5aee8cdebf1a0b5a3a50868fd6a1688336337b4e890021908692abc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/score.json": "cd26efa7dfe3e92cb52d14d8c4d41b2882c46c46ef1342a2a29b0a4fea6f14ca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/metrics.json": "fe71acdb62b5af9a27b0061bcde17d06d4911fbee2e68c6df4e5f021978bc5b4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/facts.json": "3a0ca6f5f2b20f0bd1e2feda3a92b5935bd92071df6e36c9308aa93a1e4f5b41",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T11-54-52-956Z-0dbs/codex-inline-judge.json": "3d9dfc07fc0cea65abfc3284060bf5ebdcc0091f7a1b2b7bd432e7a539cba8ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/result.json": "bd5957eee3b3813f3d1084ee567f531a6e96dea8564383db9a4941843db5e18f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/score.json": "5e1e7ba05fd78bdab471d40ed02a71c4d465ab6b6be18b336662ed5f3eae68e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/metrics.json": "b0168cfcad7363f7972d6a7e33367596db6aea6db4b06afdcd890d04f6578934",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/facts.json": "10730ecfffff73f73047e57434463e65cd3042f7f55d1ef6d8ff79ba81cb5142",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T11-57-25-352Z-ovh9/codex-inline-judge.json": "dafc0fd0e2eb292c90dab3add4b47c297f81621a0d82f7aaf4746a66f3eb5425",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/result.json": "f5543d5a0b8192b2a37159aa332ebed8b78c6eac4dec8a71c2c8d792bc0e621d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/score.json": "08008060c1e25fd9b0080e720aab389169a7cd677dfca04f2057b3c291316809",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/metrics.json": "b1cedeff981a572fb8606b7dfb0c14db697e48ee0b3be32fbf2fafb93fa1a6df",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/facts.json": "4900d0ee3fc4c1da9c518175826dfd3a3addbe6366a666b48b339b572b08a489",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T12-04-18-854Z-mfuu/codex-inline-judge.json": "0d4add27474cada7c82a18e622c270c2b6047e088ad019ff54d954453a8c66e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/result.json": "1aeac9eb68837e1dcfed8dfa3a88ba353c4149fb1b1e07bcb84f763e5002ece1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/score.json": "9dc88345c78c482c569580248760b5aa088336e20280dbd3a83c956dbd3b5ec9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/metrics.json": "b901a57b9e5840e910f85b2063e14161e811eb12186bce98290ad8bd2a8c63fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/facts.json": "4dc09037cdd48ff0f4c732514b77e290e7a6e93b2c1eed53d234df80d233bd47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T12-05-30-894Z-toop/codex-inline-judge.json": "54e6f4e3a7bd454df1741d880e1065397c3287b535ff2da66d2236cb3962260d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/result.json": "39e9231edf6fb9019cbd4dfb84f29224d770c12d2473cedd29125080b1eb37af",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/score.json": "63b4693b72bc22f90674c41adf5ae93e085af324744e25c198d441edd36e33a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/metrics.json": "eed6c5323ec340a24a2bcd84bb163c68ba7d66e74919012f9efa0464a88c4672",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/facts.json": "679eb1f70208635ffc4e57872b9f721eaf28a270cee22814a46cff66f2362865",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T12-07-13-025Z-o7vi/codex-inline-judge.json": "a37e4218b194bb8d93002b93c3b0e1918c19c025b979ff35ac0fca1f6d20810b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/result.json": "27d97f5b9617a371ae626e3a41c234e74350e421879e7f19cb15c6e0742a6b90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/score.json": "0c6bfaeda9a9cad25f526fdc3c024505c04a0b4f2751d23ff41b215c46f9f00a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/metrics.json": "ed85ca2afefae85ea72abca89fb3168a5a82a31986ee8a93d584ebf935c11d7d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/facts.json": "0bb9245e2eaa7e7c107b78edf3ce2899dde94e62c8501b131467ff1142ed6981",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-18T12-09-25-276Z-wkzd/codex-inline-judge.json": "3ebf55c71807c09dbd35388f286877b6810c4627e20a96069b25bc704907f30e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/result.json": "006dfc83108c0893da4da5814fb7a50c61f6e316a1043b9e052a8e0106d39044",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/score.json": "0053be4a7736222f844c9d29aa5a95686ba5c2a02a05e3fa33004960126fa7e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/metrics.json": "9957d2a8e545f4d6871bf356281fe71aad71ad0ed129935ca95d7376326a9417",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/facts.json": "0674668db372411dd1fbbe05c3accde2b43a6757a8a01ee8a3ab0e2077364e0b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-18T12-12-08-031Z-cuio/codex-inline-judge.json": "1b8a660d607dde579d9a2e3aeefd8136b7e7997e9d8b79e375049ff15ae948a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/result.json": "02f2a7fb643473d667b0408cb9d27f9486115443db724d34c18d7107f39948d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/score.json": "d2ec82989d19fb35c393949a04d9273bb3d8d247ff298906adb551c3ab35f4e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/metrics.json": "c05e7711951462fa1a94e4a4d483591621b3fbb425a223b65313198199b5125f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/facts.json": "101dfe8df27d6411e459b9aac6745590487141efb88cfd07b0b35eae0a51dd3f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T12-28-01-865Z-v1z1/codex-inline-judge.json": "1d0ef0421beb9e6aba85e43323a196e80f55286bc6b1df0eef71853264d67abd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/result.json": "7656492bc252a2ec9b69427f023948bbfd4bcd2afbb58a584cedcc710bb27226",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/score.json": "82d01b41f3c01a857f8ca0b0ebebd2ed2cd600a71041e218a90ef4db76eef173",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/metrics.json": "0de0a44a7afeb109e37be17002e0a3bb1f89956a2bd3bd9d0500c2daed2b28bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/facts.json": "524950117344dc588672fbff95e299da39e9133ec88a7c5cbe782af8ac017dcb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-18T12-45-08-023Z-8j8a/codex-inline-judge.json": "11c949c7356e09ebb62d58d0ea931e2be89080d4ea7cf53985ca4bd405db48fa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/result.json": "b81fce2434ec4c75ac7616686abb4dc9d5e1cc3c2991e71db34b410cd51ab650",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/score.json": "093625bc41f811c5c3ba383f392087b2b8d540d9e993f361714a32016725aae1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/metrics.json": "31c7055c9008156ee700cd80bb99c8509f106ea2086f0eb596d5fa74f6210cf9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/facts.json": "f78f832535948afd0926ba5d0c7d9bacd3cc82efea4b681f22a87e7d0104b026",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-18T12-46-50-237Z-ou8z/codex-inline-judge.json": "341092813e71ddcecb70f04c5fb506ee28d2b7e4fd21b8f307b30aaa2e4ebbdb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/result.json": "a94d1f0deed3fe99e3d2f4dc9f6cd2f247ad89cf7c4fc255afa6695b2f653900",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/score.json": "05fc0e410a70e0da05aae42ef119a58b853a14dbf68024a7160dcc3788870b9d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/metrics.json": "8325522b489e96800c5563d7355bb01e9abb3baa11ab0b30d1cdac9681d7c2d8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/facts.json": "ac63f1cdd47d3737ebc8ac7539c9f25a23a26fcda07f3f401f0a7a54f0800d04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-18T12-58-46-872Z-l52o/codex-inline-judge.json": "55352a66d45b7da2376c1aa60dd89ce3dd90076283790f9e28a71d2aa9f79b9e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/result.json": "d8e21f484c63274fc25c5e8ff3e3a049197970996c17ae316b2eae9fe71b621e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/score.json": "c6840f5c544bbc7635cb9876affa1afc63a1a050067fdab05908c759acc589d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/metrics.json": "262651e7ee3ebc3ab5507c57fc04960ad767498aa1dd523e2e204dd5e9a57385",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/facts.json": "44aa1cfcaa6da25e1d952af34e3656abb7bce5c79f2467d1408978e80c16647a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-18T13-06-08-438Z-qksj/codex-inline-judge.json": "e98fe5a7a9e60ab08cee6a42fdd7209e7e2cbf298a6c42fff14e0c97d7811fd9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/result.json": "e2bdd989303534b197b7682e2ba2b99d89770c49ea43dd250e319f88d9790151",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/score.json": "e5b570759e6402771fc207443bb2eba75b68f0cccbdb0b19c65ad251f04317e4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/metrics.json": "740f9dfcb6eff785b4fa47a039c7019d1734be0c4d8f94a1cec6043a0bb62137",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/facts.json": "657543eb90573a16a5848c05e2f7c0b6f69341d45fa609edb36822393066f7b5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-18T14-31-25-796Z-14w4/codex-inline-judge.json": "aee5dc4706384eb3b1daa28e94245d65a07063cf4503a0d9afa37d89c5cc67c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/result.json": "fe77646186ddb71839f6aac56e3cdb6ab7b16819924a1dec71729908790bf604",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/score.json": "cbe328a4af419f54f4b96c994ce22772a561e3a8f8ba9a5b5aeebfc213cd084b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/metrics.json": "eb139c0705ddcc4654048cef8369768a93c5f0e328e9baadc6d54f6b4be0368f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/facts.json": "1f6ff83ed902c39d35e61b2470c605529db24d7f17aecc7cbe2de4de793120e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-18T14-51-30-330Z-rv0d/codex-inline-judge.json": "eb952f389b24e30de791eb32febfabbc14bcecd91ee45aba18475e4b0b237367",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/result.json": "2ae3fda0b310aabf92d4631d19c67befba12817409e13bdf29e6165a8e9dd868",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/score.json": "20d5c7b8f1cbff472888f6d0affe1953d0fc11c5f21e2efa36163ccd1dd7db0a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/metrics.json": "63b74bef07195b60a78cb6916fe29b274fc91c517b45338e7d3e33a031bcfaee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/facts.json": "f63743fb7480b49b285eab1c5b4bb13603fb4ef32992c3f687d3dad8efd690cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-18T14-58-25-853Z-aps7/codex-inline-judge.json": "9639364e26e9a926cbdeb9372617404f66f475032d54f7c914c0a8db3a838ffa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/result.json": "8c4ccb1a2d647b82bf0fc2f14b49d7ac5864bd10c53b09d2c444af3b79de6c2d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/score.json": "33a86e8d3dbb0ecc85802f0189f53be36dca072d98884c3f597a141185fd71f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/metrics.json": "fc39a801bfda17e175073517e4edac8f0e662310e6fdd3e1974c9e9f209f6901",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/facts.json": "e69fd05275447b657ccda6baf6a0864b476a54d19717c53d80cbacfb31bd64f2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-18T15-05-14-344Z-21rr/codex-inline-judge.json": "1056ed1b87d5bae6bc5ee8ff06c4a92cec01cf3f972f5c22ad3525e8353bf762",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/result.json": "c3486286e37e01f51afba9a7c04c63dc0dec2b8abd9ff0de7c3c41f59d9f7649",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/score.json": "d0672b184a9ddfefe3e9ca08574499c9169b4c9ba4f2b1f8abc548e902082526",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/metrics.json": "7d19f95a40d2e724838c2a10cc8deb722d70b74af456932900b31d67cbd56583",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/facts.json": "327cde3d647c5322945043cac4950d66befd8516c686291f56fecf8d6c4c4e2a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-18T15-07-36-652Z-nzav/codex-inline-judge.json": "d097243a74756ab17b4f15038972392e24b248096f1a2d5bf0665cb229ffb44f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/result.json": "334fcd5c43cf731393caa89759d190ee00062f0df0c3beb1c61200e21f287a8a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/score.json": "8c39c1996223c0101ab5803e73b71de4c412b212d24aaea9c23583a87f532937",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/metrics.json": "b27e272ebd0e4cba848f29c626c99a291f028c8efa656d80ebef7b35e02415bd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/facts.json": "c4f2e80aa6c2512c32a822f2a94ee1e512a075968db0325d9365011c8f2ff6ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-18T15-09-03-717Z-kxas/codex-inline-judge.json": "4bbb6a7d8cb3a6352e1d8fd2e3676bb41d253af09767ec7a7a8e39ec85c1453e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/result.json": "b5251702b4df92183d9fccb5bf971961e5cb2a1646f8d51750e547d9e92049db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/score.json": "c2d8f30e7e88344a06c2c3e025db6d4a4ba49433de8510c66f3c51428fa0b22a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/metrics.json": "c45ad422e1b38b10cdad9e8bfaad9710fc4e47f08066e5418ee5b575d75d5aab",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/facts.json": "7d2ed117ee5a604f6515200a453f746fc675c5c98bbd3f7fd12d5d815bddb732",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-18T15-14-46-925Z-x214/codex-inline-judge.json": "137a4975d22c02e0e1051f963d1b1034673a01c4d3d3d922034639ba7fd84d99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/result.json": "90374d3c292216509a1cf3c62b07aed2bb13b0c6de62801dfefdd5a380dcf23e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/score.json": "da66c3476804e95dbacd25ee3bde4bb19eb2c50359845b2aefb054b28497a87e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/metrics.json": "901be05b86dbd22e8bb0906f73e758859f64da16ad805666543d34073634dc53",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/facts.json": "5b16ae1f7d0c210d16f19f764a3dcf46b4e4605b7c6d519ed454402da5f21fd5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-18T15-20-42-276Z-wwm5/codex-inline-judge.json": "b1d42bb2d65b60184613cde9288a6300325aa0cb78dd0780e6daf7877ea84eee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/result.json": "186196b2184d206068f0873e049cf31902aee631a303167ba1eaa0fb7a8bba90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/score.json": "b3014c3ad5f538200230547290691414e1f67f2cf5f0daee83275b786444793c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/metrics.json": "11a6a5a5b880cf1392202b2e47370f5e20b84cc71f351eb4f0d3ac7b9eec8a38",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/facts.json": "2b10a65deadcbcf434bf66b38487b653aedb50bc0562df4e252d8619195a2fa4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-18T15-38-06-057Z-ckq1/codex-inline-judge.json": "1545fdaf556f75db77872002bd884dc9fef6aa2cbf64402cc5594296607ad1cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/result.json": "adbc73d2b68f6e97f3ae717396f13cc823b0d1e824bfeca79dc6775110f0806d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/score.json": "f2e0aa5598e35a68ad8f81b45398e0da9b0d834f01ebd66718f1ea9368bd5340",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/metrics.json": "645045bce3256ea7e9ae337bf9d223daad6db8d8dac644da8f572ce0e44241ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/facts.json": "2688ea75102275948e6c2c7272209e7f718392832365d24c860b42a90032e041",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-18T15-45-32-686Z-d1t6/codex-inline-judge.json": "476f39a891fc904fb09a3203021b1bb315c465898d81f6700b2da3b3927c067d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/result.json": "6a6d68131450ae949f074ff5849196e27391821ec6c96196813534326ad24f28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/score.json": "c37c92d5b16bf975bcbad586948203be3d19db72b500735a8fe2f3d57ace061d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/metrics.json": "62bc2374d4420f921ecdc5af2e5e95f78046ee07d1e9cee8ab625b630326af35",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/facts.json": "0a99e255fc6a33546119607d13c6436e880ada0e1cea5cd0652555ac910f6b83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-18T17-03-43-989Z-m124/codex-inline-judge.json": "d4070d8d547112221365d9363da58c0be8c2288991cf7a179af563cacb8795e9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/result.json": "f1780647198dbb82e425dbee4322bd98e0da825b19b449954a2656519248d8a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/score.json": "1d2ebfd3d1423f4b8af7995d05c46c739cacf405ad0025b101cebe96a3d9a2c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/metrics.json": "a61e3a71e95b372d173b7db5e4ed45fd56277f00d8f2441a88df9a9336357995",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/facts.json": "ab93344e88988e3f417c1a1f7ec0d9e0d9fb2e62558e95f3b93837679d711979",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-18T17-19-17-637Z-4vdw/codex-inline-judge.json": "d506efb3408067f73ebdea3c8941a7775ea379c2e8cb912a998f10a26a833e68",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/result.json": "48f434cae9231881c32315d5981f2fd865e8ed2f08783c8832d4752a61ce11e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/score.json": "5e9002b49bfb58976e981cc988c94a3293621aa0a1ac4984c3303cb6ba3859d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/metrics.json": "ab51f297af2b1e098a9b9fc4786b50044c0cfefca7bbd368a4c88308e4a3cf96",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/facts.json": "ea146effd5413489abc2fee221dffb3e76b5cd6c64999a06650751031e1dc7a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-18T17-59-57-472Z-4p8x/codex-inline-judge.json": "882a45b07add99fd9dfbfd057fb1d0ebc12041b390713b18398ffafb4cf043da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/result.json": "80a7162a003a0702438dac4490663583aaac1ed95baaa60b8bf65035c939b671",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/score.json": "002464250978c61ec12ebebaf2c4ccc41eecbd42a5723f1e6eb2bd6d5121d3c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/metrics.json": "2db6a7bfcd61d1c68cea2f012091fa34714244ac9ab3f66bf1015517b70f3b42",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/facts.json": "8e4befa3afdb43413bf732bb1b382764c99e72dc83000d9341b8ccaf04180bc6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-18T18-09-59-804Z-q7nn/codex-inline-judge.json": "8216a25357420aa99660802ff5808c41038f2506cce1a683e96003d3bb19e436",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/result.json": "0d95086bbda9d61164bd4965cd6f56ce40928bcc30eeee840f03251b3ae296c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/score.json": "7e6f2cbe8f990ad2184a87ced1fa98df18843131b998bb5f940ba9cd68e6757d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/metrics.json": "b204880b482bfaa6ea013d4b025c3ca5902cfad53b068c7e526dc75b20ba2bb2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/facts.json": "fc6cd5ea9b25acd71c975d351e005adc7133ed471bad925bc26cdebedf8ab241",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-18T18-11-14-804Z-xh2u/codex-inline-judge.json": "d06236071e0be4cec0697ea426a1981268a1030a5707dad09e091ce8794820f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/result.json": "52f3215875a4bfd35aa0188c5a8b4d918555e2d4cbf67178ec3f17b99360f6b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/score.json": "b4166da23cf68d4ff8acd7eae971f46a5d36c2b4f65e5fcefc0b37b9393a0f9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/metrics.json": "8b84bf806c4271a764c692b055961a3dd1abb096811965ce4a0b70c4b01126fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/facts.json": "58057460d453db2ff092b7de174bdfddeeb5c4bbc84320b771799f12aafbd0cb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-18T18-39-04-688Z-562i/codex-inline-judge.json": "e2954db2bd9e2fd4a498242b700423eb7b1e767ea4fb3dab3f3cb72ca2e6fac2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/result.json": "489ad451ac99672ed9da53fb4b7584a7a007a896c8bdddd17889f293649f58f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/score.json": "fbdf05ecbf2462fb5a955263c5956f70c867698c62eb7f858aa80be12b056ef0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/metrics.json": "f22b558bd65590639614603ed2af6b6cd603a9faae4b77f97e6f2c72a7b724b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/facts.json": "de4da16632cf42968ccc970618bc4fb3930f291950ea33a1aef6a1027bcd1113",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-18T18-49-55-293Z-d38x/codex-inline-judge.json": "fea846f1a7a944fc3c428b85a1cf6b43895338e775e8950ac13dd8e57033d5e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/result.json": "b09d3212d87011b3450b53817eae8ea180abb53bbf167a770e7a87819b0d9ece",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/score.json": "aca4c34c915a912fb39bc0baa2f78ae01a323d49024a6bcc38ea28ce86f1c5aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/metrics.json": "61f7ea7a52cb9481602484675a7c58f2704009e259196a6faa395d07620a0c4b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/facts.json": "0a4bcb2b477c75023cbd8979c51571e5e1992374d310ee99312c7bfa1ccafd3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-18T19-46-23-504Z-hmqo/codex-inline-judge.json": "f029c5e4d42da6079fa180071eea1bd5ebae107ac6878668ba0f00b1d86ff405",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/result.json": "0a204db9089d4e948a36f3961f2a2aab78669d21b87772b22bb371594ef64634",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/score.json": "af18d44532e0bf92e5dd46b8e5253aa65391003808e94b1d97450614ea69fa83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/metrics.json": "e09556705a325ab86e5479574ade657a39d7f8bb9a22691a6be1ff42b8448988",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/facts.json": "cfaab2d076978c59276ed420cc54517f03527b429d28e6db89db94ad437e17ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-18T20-19-29-433Z-tu64/codex-inline-judge.json": "649ea1eb11f0014528e4bfb6e10c0d3953a22a55e3a75e4657ab9307a03a27a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/result.json": "3b91932358831d39619a098fd3aa5a9b7668b17486d53345ae59fe2231221d83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/score.json": "ca84536a6f40f43e8fd5ae218b4a98ece94bd5c7345bedf52de54d7c6e1cfe83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/metrics.json": "608e8410a1c780ada8d50b62002d031928858e5e85efe80fee59fcfc2638328c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/facts.json": "95c23ccfa77a26e60468c1ce8a67b1ed8d6038c6557cd6c8a806437255fb04a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-18T20-41-27-919Z-xafd/codex-inline-judge.json": "f09bac16f738a45956ed85754affd8abd78f2b967f85ced7046df2c9f6233a2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/result.json": "b3852e8e0ae8c407a5240394fad07eeb74fd35185e54111fea643a85e82c22e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/score.json": "ea8b6cded834d864932b7c7c42b43fbdcea3cbfa4a72ab6c7ed5838a6fe46925",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/metrics.json": "77436bca24b102bbe5b20f1e6d54e967ae4b14787fd4c79fd1b3be4145994602",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/facts.json": "4e59d8de03b07991b0483c3d3bd76575037f269d1abfde7358e7ef55c8ab8cf6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-18T21-25-22-821Z-vhj3/codex-inline-judge.json": "b11a3a97996ccbba0fede05e768a3d6f968c82daa4a3e210d62ae66b380a058a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/result.json": "f97162e46a0b5161a2254e758563440f56cb6ae23da279219b1ca9773ad1e48e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/score.json": "c931a9c913fa1b64ac38c1f66dbd5f40065bcd6c4ed30ae9775f340215d6bf1c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/metrics.json": "808ff10cb23b9e4ceedf75a799c4d06a3fcbe562f9c82f93726bf04a1927ff4d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/facts.json": "fb5b599f16c5adbc0bee9eb19ae0d2770c31606d23db5904b3bb5bed24cdb2cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r1-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-18T21-47-11-728Z-n5za/codex-inline-judge.json": "570dc4ba1451bb6333f4869f4d4e73a913823af2ba7fc6c6992a8ddd22adefc4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/result.json": "e790ef0d513cf1ded86ebebc42db407f13a43745d27b50a41394d6121ade0195",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/score.json": "cc643953abbb72f234aa1982c3633fef281e72ffd57fc1389418ce65885249e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/metrics.json": "bf272cda6dafd3fcdff84ec454a2b2892b25457db6136676b96d9d7ab5f688d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/facts.json": "46dd3668db6885fd8addb22478ddc5a1a4428d8be0a433b17d0aee327b114ad9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-18T23-10-36-972Z-qdmj/codex-inline-judge.json": "0e5e9cf539012807bcd42f4bc801d8e9fc8978fe59ef2f1cbba7d80a139375a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/result.json": "8056594ed41674354968c1a2b8ca98a0fba6f3f8df25c2270423ce5b6ed8adb7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/score.json": "adef5c8f21d81bc84145dc469d556eb536992fe3414ccced09442017bc66dc4c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/metrics.json": "b15f1e2c9afdc5633f1383e9342f2ab366de46ce0ef2a3aba761db96a95031c4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/facts.json": "54a5cb5a381a5a04c2576f2fa7396b7eee211f7c921bc2d12cbcaad4ed38dee7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-18T23-14-02-857Z-avw4/codex-inline-judge.json": "107dcbc962a5263ed7af83f244c82ddfb344f879fbb24700442cde7e71f16903",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/result.json": "4707e019eb773939d97f611783672f5e9ed9214a746ce3858f32523d6269ed16",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/score.json": "f10caf58464d2e6dc485236a0e62ef4d9c3e7b86e2381b8825460f8e934138eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/metrics.json": "eafced40875dde7b227d7526cbde74db91c2fbfacdd99547e9d1e12196056354",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/facts.json": "47467bfbdb511dfcd79e53e0d64abbe66efdd480a9076332f71b4abffda81a0d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-18T23-17-58-672Z-wt7c/codex-inline-judge.json": "8c0da268f7734059ece83547ba49f435009187f0c627c92b6944de71882f904f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/result.json": "ede323501af0b1267cdc5fae043c442af83ce84bd68d5be73c73e7ca990d994b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/score.json": "653f633c9d11c549c8fd4d2b56ff74c79a36c5349d2cab1f024e231895542656",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/metrics.json": "a354f63bf88b9e9c32793bf88ab3c4747787ffd685f6cbe571c435e9a6baf016",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/facts.json": "7a60d946645a35ec9f746c3b45055a3e8c172f1eb16da3024b85cd7323cea9e4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-18T23-22-00-557Z-41y0/codex-inline-judge.json": "d4f4b8f0fd81b9fe9a017f664459460765db84edee366ab04a3f2e11e901ed26",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/result.json": "10367b8765bdda7821cbe7138da246b2b98347547da3a746855530f1a6d20672",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/score.json": "2300a182780e5bd70e7bbcb43ebd70c2a414a4ad762bb91a6d15ad7a44d2c6ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/metrics.json": "dfb252d0320c8ea7e44265079bdf419eb8a738e67cf4bf71142d21b905c7b60a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/facts.json": "17f5360208c8050efdda286985eff2bfd8685241e70c6631681a9e55c99354de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-18T23-26-09-472Z-tv8a/codex-inline-judge.json": "93b31656bd7970de37ff07cc7cfe63f99be2853388c00f2dc736092f5e53ef6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/result.json": "9429b07a98696e0a5edd05ca14f8106136438cec0df94428135e5cb4232645f9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/score.json": "30b6286caa748f76dc0bff4ca48b7b173384ac45ebeb23b936ebd536c9881f1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/metrics.json": "e05ac292c4c8a8e999da1a74f88e98bdc4deed4633aceffce25ffce43d762534",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/facts.json": "1b19903106209e4b17cef8391147c8d3be1da86715be815e9e4f2f7c52a661ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-18T23-27-11-491Z-4p66/codex-inline-judge.json": "6708b9c869c96aa5f7a43d64843faf9e626a52c77f5f4c377556e341e1027deb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/result.json": "b477f7d83dd5b02ed0d8f1a6b3ccba536c71e3c7b6c834d889a5ebe1f2525581",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/score.json": "fb56f98dde87446d11724474a2ba05d8586d743c5e18ab4f17b45de8a7a3a589",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/metrics.json": "ce310e82f15000959fe01240cc8f27191f145231bfe3053b605c31b2bce58058",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/facts.json": "bf5347082711ddd9eba187321883f0bd410f994c178e032161bc444b06473e63",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-18T23-28-03-443Z-yieo/codex-inline-judge.json": "c28936cddce7a44b8c6ba2ef81444586adcdb4a46e4c81e24e77d6a64e54a74a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/result.json": "b54e670ed9a006f21fbd28fdcff84de59f3695e93eb29bda1fb918e67efcae09",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/score.json": "717b9679357b978ae7c1fd5eaa93012c772b520c24d3f0a67d2226af9847246a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/metrics.json": "3b1f49ba6fccb7e9685cb883d211fe2aef1e5fbdcacedfc9ee39a91424705b2d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/facts.json": "3a718d9de1de2406836678141c2996caa6aa863cc784c21245a2d2ad40d95f05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-18T23-29-50-613Z-tcg8/codex-inline-judge.json": "f5f9e6ace12efed6a19ffab3f77fa2cc25b1c05cbb61faef86de13bc23fd58b4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/result.json": "61dd15562f1d29beaebba6b73e6075b6c5aec9ecace7c681fd7875ff827c7616",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/score.json": "41162502993d708e9aa719414bbda520d6b549e48f851c5ea080d6d6c273c58a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/metrics.json": "ef0b35357c75818dd9fc7e7570e79aeb65f98076425305862fc7d65dbe193170",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/facts.json": "b49fed6e974d8fc80547e03eca9a65de75d62d7a1356a4c31a0c1377c7c911a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-18T23-40-40-174Z-vw3t/codex-inline-judge.json": "502f25430f5c35487727c690ff42808372ca1be7954fe284d98b6caa0cbf6a2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/result.json": "7e60dfd8b9522ac06cd38541f71cbba3e376f256b4b2066d6214d7d7655fc037",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/score.json": "ec7b0da2a2d7b5992cef388d0647fc913cc3acecf35e0f6174446a32d3fe884c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/metrics.json": "803b5a0bb5beaf8fb7f4665fa6116e004b903370f3e6b41d591dc90e8b136fc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/facts.json": "98f3a79ee12dbf7d6478c97254e3d30f0cca8aa631527f329d483c5f110cacd3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-18T23-42-47-488Z-tu0m/codex-inline-judge.json": "555b15562cf511e8f520f74c1e67ab80b90325b462a40c0f0d0e6c461461f68f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/result.json": "ebe2a97925f13b6552afe3f8f31ba48eb399a19e2f0aae3eb2f6af546a985ea0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/score.json": "7f41769ff94d78e49c6e1f7ff14b8d96a192d5a10b4dece8e0477439d35b43c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/metrics.json": "3cf6c6f78ca8c8322181ac56ba69798ae309c982e6adc01ecd6ffaa808d22b0f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/facts.json": "13132c0308510ca7ed597eac43269637855f909bd8c4a3c17c33f97f51e73a1f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-18T23-45-19-891Z-eulw/codex-inline-judge.json": "2247ce2a1528b987b7eb138b602bb3d8b1bacaaf66750f8dc9e5c72e420e1189",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/result.json": "4474393851ac186e85ecfa90aa20aebad9ab6a5c7ff17f76298f5566cbf4aec2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/score.json": "c00f4a36dd1264c6b52f51daee52bdf5ff887d8b0cb519406383014d51888753",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/metrics.json": "4d4545575895d36fa5503710ebb94201edf3c3b6f9954a7aff331cbe060b8428",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/facts.json": "d0c0cff3b4dcd364b8c045960fa4fb342e1ffdca9092c4de7a7c42e974d87f24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-18T23-50-42-989Z-1yut/codex-inline-judge.json": "62d1cad2c3d5318a862179bb3b714248082cc31429729370d22693dfea77b5a3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/result.json": "996577ec37a79ec4c1baebbfa4c814f0f56c4bacfdcddc4f377a986de89423bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/score.json": "bb85b1f49e1cd4a065ff674e06d36943be91101656fffbb2a3490b987a8c5db7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/metrics.json": "7f52cda5c6510d3d23af60d8cc7568cc593be6748316a9ba12f6899ca1079641",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/facts.json": "e27f2c528af0fc7e798aa4085722b2da792e42d44ba9559ceb083ff71d4b895d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-18T23-52-20-246Z-by68/codex-inline-judge.json": "a3fb840a6f4d5ffc10bb9e74581edc57d75aa5d98dfe024d2293ddfdc78d6245",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/result.json": "bab3e78b1ebd81faf567a684bb846f66e7477b9c8d2a5f9da53013778605a14d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/score.json": "6d4674112b448dfd5fe9ad9294e182eff11d18e3471955a4993c6bccebd0a43b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/metrics.json": "a1835f67635e335214c1360ae443302fbd504416240bb66a4d690cad0ba09927",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/facts.json": "848cfce1e4ead142e8225cc5bf19d07396a601974d38a75a0c5c6e9319579aa6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T00-02-55-715Z-qvc4/codex-inline-judge.json": "1febdc7d34ce2e8f8d510a6da8a599f8f91d4cfccbccb2df28590759cf5bd369",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/result.json": "614d1e56611b4b7eef380650d3f62c0c875db41217f09a2b5e261a3e74243f78",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/score.json": "b8d4fc687fa0623c585f61f145cee083c21b70439454a3a38a48c09781c7bc6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/metrics.json": "bc74aa62486d4397001f88d955e981e8960df0d00eb9711914752a7e53aa9615",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/facts.json": "32c0ec6b350d4237693e499e5c83af8979206f83179027d8b31f2b4de16d0bb7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T00-06-03-261Z-wgnz/codex-inline-judge.json": "70b9b81625a5989c2c965ee0815bd9887a81934c9118e51f23cc4fe189eaa124",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/result.json": "02fe3d85d8fd601a8fca1ed2733df312481bda6c10e2cce5a45b10ce56c0b38a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/score.json": "9e291593d4ca140b65b9f63c446cb9883e3df252377efbbd28e0a9ee0de6acd5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/metrics.json": "64c9b60a3d2cdacb954470dff8b63921bfc4de407f19b8eb97f4b590dc5d12b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/facts.json": "c1501352940a4f8889cf4e633d6e8de862f5361464bb88ac81f66daaf5bdf38d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T00-08-51-051Z-51hh/codex-inline-judge.json": "8428ff433785f3611c50dd642b0df8f97118a362f6235a4af021cee3238b780c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/result.json": "44bef212419475b4f67f1fb41fb1ad59a6482148d9d85aacf8c4508e8bb13b41",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/score.json": "a256cab4ed17e4779349a7d16445ad4227fc8eb13693b32293ddda8dd551073f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/metrics.json": "76a9bc7f0b3cd27e8ff36053b8f6e5598b7bf90743097090f9a15ab7b7e77760",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/facts.json": "34a4968b7efd027249a65b2e60caa1996a0aea48b8065f9643aac5926c98a2ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T00-26-17-349Z-8fix/codex-inline-judge.json": "b2c5d65c4d81f098491715b437d23ecd74c31dad8f2e884058c90dab0ad91989",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/result.json": "2d24dd6977654c4c6dfe8593398c9eddb5680f4d0649875495ccaeef88c8d15d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/score.json": "5cf9301bca24449d59ebeafb0b7adc400b3ce874071c09b00efa1687f9d8aa99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/metrics.json": "87e9f8fb47b091b53cca5511bb424d590b439af3923f230d69ba89a5724aae6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/facts.json": "478ebcbf5108faaf16a7f68e68c3b68d4001e87323aeb3fb0845c0893334c3a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T00-32-55-783Z-fh89/codex-inline-judge.json": "37cae37b8d018795d2100d3cc320029873bb651e43294bad0794d90fc8529cff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/result.json": "852c6d6b0239a37c0a01379039d1512cbfbfed0f75f241423b181f5fca62a215",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/score.json": "4b281bcaa83c21873c814c5243443002a513ee8aad80f86b60ec412781e24502",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/metrics.json": "28be6916c4fc57a67c092b2446299717bf1ea53442372dfbd223f04c7c289faa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/facts.json": "787035f9d1c2f9b3e4ee7ced831ba17b53def4e6e97d9d57117c1bc3ea3e58b2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T00-34-48-043Z-iqem/codex-inline-judge.json": "30bb80ab0faac41c968da8d96f24142d259501de6b53f608dde43ce8ecfd5aec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/result.json": "380d4eda703333a1f304c87f2e4b078b69ad220d2121508e345ecfeb9436d1a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/score.json": "3740412428cfe679d938af54f2c4411cf5e254edae4ec2329243b0b99927d517",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/metrics.json": "d81f68baee7c23de2858a8d12c713adcc37f3900c2c0cce0e2e7d487389e1bc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/facts.json": "c9b03d03be7aec3699114246d9442262115e4cc13358c957d3271f57d5833857",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T00-49-09-635Z-mmip/codex-inline-judge.json": "dc43671cb0b7389f21bb479f07076780f037160aeb619796b77840202818be8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/result.json": "9c2d89b00a8ca4a92b96c3126cdf7e2ee356349b62def7b1df733bf0b59b8a54",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/score.json": "cd8a193ad7f7c3537ef7f3662e4db55ef6f0b0dfde72b64d9cb74d9a36bf2c8b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/metrics.json": "f53c22a89280ba7a6b31d972030f35526234656adf5053448c2b5d29f00b53a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/facts.json": "59e27f26ab67ff6ab1d50df3eb94d14cd301f8a46ed4da33c174b1fbae361e6b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T01-00-34-340Z-qx0v/codex-inline-judge.json": "8b9366eddbe11dc8ad642237e2b427eb66d3e914cfa5df3993f2862e2a81d6b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/result.json": "c7675fda31c6b27ce73eb5a5acf78ee6f1dacb2bf4bc0a3a67c50838de20ff7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/score.json": "e667c72bacb7c61340af5d7d5479b811249863bc1c123efc74fb84513b8deda6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/metrics.json": "a261692756df46fb17917898041e30bd2f66df92288e9e2f0e12bc317fb3a905",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/facts.json": "01355454cd6b12e3e289f4e1197c0d22c0e2f5d2538be90b0c90c54aaf236833",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T02-30-54-417Z-ogri/codex-inline-judge.json": "877a268930277039d2be245803fc7a89b6796a7df833fa63cad6e6fe4a741fab",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/result.json": "d3e6d29bfa8d285e458e6e150b0b543b6ee80130b562389108eb1e1149c307f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/score.json": "201e7ef3e824bebd9982ff3070db56c2966673fb639fb902bb6ed28abf7724f9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/metrics.json": "3e1a88cfa9233e8279d43a4e0cdda14c9a263301ae106c9aee081944a327380c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/facts.json": "9a5ec496773717633e0c9e6aaf14e869f2c37cef55f3120710edffd5f1d97a36",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T03-15-29-761Z-kro1/codex-inline-judge.json": "6d9f6c131208f656aad33677f1abc82b9e4243d2209e4c87c6805b019bb8740d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/result.json": "175013090a73b35776096f88e079ee890b8d1dda15098aee20b3f3772309d52b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/score.json": "342096225cf0c58b595c80456462590c54b0984f0fec887333a2027fab4755a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/metrics.json": "4593b751e65ef99c0c4408e8ae8fe45da6827d48196e19a144426bf573af4b1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/facts.json": "95317f71ec685972b0f8659062d0bcadc0b7789706a9e38c197ad52e03e09132",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T03-24-05-689Z-yrbi/codex-inline-judge.json": "f3e93c5f7eafe2f3162c68b81587621817959c65499b195eb666c9f920106b30",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/result.json": "04b5a0d1e1473f6a37c5eb5bccba4b3e1ace9f17fe898e382c9645312368e67d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/score.json": "a2de6938a1475733414e498dbb2983f507432f52f3a48aea1c778efe44ccd50c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/metrics.json": "e52af4f5397b7cb7d6aea7e35b13f47e413fcc68e8681df7f082130ef332069e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/facts.json": "42e7942dd614a6e331a70283681a19856088ae5ab4ecca84e75fab8cbcd9d1b3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T03-33-31-871Z-yyol/codex-inline-judge.json": "bd8beb6ad8ccb9f61a6fc9ad382ab2582c3e14ca5956d4fdbc7668a7ecd2a332",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/result.json": "55f371320c9e71297be9813ca14add251406c8d6ee90557b796cce40c1bcbce0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/score.json": "d119b8a2d95f5290724aa83bcf95057b8d4c7ed35229050af4aa01eca38a6d53",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/metrics.json": "e6056ded72668e97c74a393cd6ff32d85e45e8b6759eafda503aeb3744b8bb93",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/facts.json": "b35dba05d0e25ff3be4f1343ee40bd71953b707804d42cac8ee54bba0f295c4a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T03-36-01-231Z-mzi4/codex-inline-judge.json": "94047c1d472a4491d37124c149af6ae61db32895d21b7a76abf8a74e36cdc9e8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/result.json": "d7f3a917c97248a1b0de1106704d0876d0ac7ce114a977eaaefd65a026e0a643",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/score.json": "e6cd6a945395d21769fdbae40c12fbd8834754eac6bddc1c026c2ee0f34672e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/metrics.json": "fa4013d8d23040672d3f9440ebebbf934b475d7e7b58ace04ac9a9e72dd79dc0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/facts.json": "767ddbfe3dae90d2ca7e1e81e3af168d775f361e2e2eaa66a602d61b17a887ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T03-37-28-341Z-y9hx/codex-inline-judge.json": "50770cc227f4e39da798c63c72a42c253446d8e23b198f5658b13567cc829158",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/result.json": "57f6dfeaa918b3820d72cd8765fb640f2a2db6c94e7bfb216e290c258647d059",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/score.json": "a8927ff87330e7273f7e88cff7138d4064d5350580a0ad56798250170b27ebcb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/metrics.json": "b267adfb87e3674325a4db12978a027bba8e2bd7647144d253e7a81c28374e8e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/facts.json": "47dbf1cdc0aad4c6fca63baf4ee546596814f970a729789892cfae3a81e099cb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T03-39-35-563Z-5667/codex-inline-judge.json": "6caebe00c4d984d597a713fde4b32242d39d260c940aff245ab8164d7e3bd5dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/result.json": "84c3a1cc0ac0b1d70cdfa599d53b874fed9b305e0300105031fe8cc02d8783cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/score.json": "abead70afe1e352ed649ec9d4ac6652ebda92a34d921af09824782cd86976069",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/metrics.json": "5b5d27a56eaa73bbc36f2381a3e635002596030c8ff62e79caa9a995d392515f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/facts.json": "22c782a880c230a5e84014b6d461a5af3c1989c407f40f0610b6075f694d6f43",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T03-43-06-200Z-n1pe/codex-inline-judge.json": "84858b4d70fba62dba53b90b5db7886c3db62a07dd768c212afc373c59c4d886",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/result.json": "646032194b08af2afd04bd7d4369f1bc9c048601bb108c807b32985d7ee601ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/score.json": "ec98dc5d3a2fceb647c33618390c1292d8dbe4b523b014741513748614d5bf6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/metrics.json": "95e79e9c2377237368bd20819b1ba8a959e68509df36b4db0ed59ba9da316656",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/facts.json": "06fc9b0dda306411f7d4b2b181d6153b6746f5f3893b87fec8e76053500f3e6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T04-15-02-944Z-e6mn/codex-inline-judge.json": "7137f9046cfbdfc3d71ac06b2c924aa4e0bc0d08e92931a49aa672a421ae2a81",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/result.json": "dc26ce26ac14eb2b3d7939ebae414a466b982608ac40d8ba32b13afa6fdd02cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/score.json": "95cd2252966704470b3af66efe9ba4b0819ab72698dbe5ce5e590bf1bfa0ea23",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/metrics.json": "28d0311bfbe62fa7af36b270449b8bcf33a612caf69b1c11825de0b2be05ab65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/facts.json": "7fee9507d8c63285b7925d129cb0f74364504176a0010106007be83fdd506173",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T04-21-03-208Z-q862/codex-inline-judge.json": "38620d764c31440eb6f0049ea73c3522415f98004767a0f911b5ca134efd9e0b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/result.json": "24f251661321f9b7235c039d48b597d9ef7e688bdf005320c821cf8ac38c1cde",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/score.json": "6dd46c2aef3e55182b7e738d88356972ae1e80609a2514e70b40884d372a1682",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/metrics.json": "a2fa1be50f6acc89af8a21e245336780477d731d619ef247314c5152a5475496",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/facts.json": "4f5d9a348f29876da267bc492ffbfaed12b56b93524149a0b3672c0b1ac24d6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T05-04-57-176Z-n7dc/codex-inline-judge.json": "ce0abb87ed9226c92ec4c9f4a3108e99eb7e9b8bda253ed4b5b59055c1d5ad57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/result.json": "4b0e51e77a166bcd71f1e46aa1a7c9298481b7f249f8b7b77e0660ae4a5930b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/score.json": "19457bf27e012106fecc9036797050c38b9e90cf955370430b2490e982451382",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/metrics.json": "c6c2687ac80c7c1fe8af78aa5913ab0c1fe1286c4ef3f720151b40ba781d22f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/facts.json": "2f57b8f6bd3974daf0882f38e9bd4223cd78fdaee1c3e0f66e584ff3a98d8cd5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T05-18-50-472Z-tqek/codex-inline-judge.json": "23f9600fde83eba5ba88e4a26360e65a4742cfbe1192c297ba7ed8bfdba4825c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/result.json": "7f57f9338cc40fc68e77f9d9906fe07b72ee325cb485916e06927734af8a43ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/score.json": "873d869d39f674044e230d865fe8ece10bd7d62761f6a00ea98e0b762ac9c645",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/metrics.json": "93a023fd77102400272533e3aa4cbe5f94f8f2218cdc3c25c59ece4f2fb3206c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/facts.json": "b9863df872aac115bba6edf5a014ecb90c4ea4b32fdade4272e22cd43c65b9b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T05-55-16-429Z-b8dv/codex-inline-judge.json": "3c91871d91e6fd25a87850a8f5ef70438fbc1ebce5bbc57a5f433d21de1ec12b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/result.json": "00ab97655cf83b4f8ff22837f2e91f25f7dfa6fade558c676ff5ba9b74a3fabe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/score.json": "e645c10eaa2b5a1f9798a878e7452a8196650e46b656b29c31b4830b2869494f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/metrics.json": "a6706fa0f2cdbc103b2c78224cd81b3a6adef2e464b80284d2a372d7f25f2999",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/facts.json": "ecc9f16269d6119f09060c45957e7957ee0b9c61b4db60ecb0702e059f5a83d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T06-04-52-685Z-dwur/codex-inline-judge.json": "b88bc315c92386668097cbf482fff0bdc3ead0a896a69b07991806759b7278d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/result.json": "20ef67880540eb7653ed80e509118561a45845ccd362a810a2065c26f3b1cb75",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/score.json": "06827dc66dd01438224d984b24d55fb18d35e53f72aeb939bd90c755bcb87368",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/metrics.json": "44b17b534679941e63521296e4fd8171425fd34560255c610c2197d736739a1e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/facts.json": "24c14d1a5d0bf8f7e82b19c29b9afaaef01dd8b68806d812ef3c4518295c1e2d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T06-06-12-771Z-1y81/codex-inline-judge.json": "710b08bccde45448c78a2b8917c2cc297026951cfe9119b3e6af373144395496",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/result.json": "047414fd837524caafdb1910943781ed123ed44a0bbee9458fe3e6b9f7a518ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/score.json": "450a53fe03704717f51ea451b25ed0ce50428873994d67802704756809868aa5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/metrics.json": "0977350c9219d2418ec92e7fd1249e8acd1ed8228dbf4029d946ab6ec4e0f268",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/facts.json": "b72a90065943571a3539b54e63d6d803096d5c5714d29a43244c78479287b559",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T06-24-14-204Z-01bs/codex-inline-judge.json": "6e2aeb5a895c58f2fe18b87ec02ea9b3bde40f8901d45d4bf108adf4b1b83092",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/result.json": "51a73061777b0186104c17d9bd902b4dd071481f950e746b6a4820b5f516410b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/score.json": "1b3af60e83cbd9ed345028290d7076b5aeed2cfba0e49f94d97738b97fdb195c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/metrics.json": "fd7dc9eb889dab0c16c6abed21adecf27a9aadaf7e54bf3c4ccfed0b719a6ac4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/facts.json": "28f2a140f3a882fdc1c02d7bd3d0234679e73539734b37a3dfe7f916ab609d3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T06-34-46-696Z-i6pp/codex-inline-judge.json": "bff32c5d1df9122984bc161067dcc1a098672d1ca82d2a4917f8d3b1dfac4779",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/result.json": "fe58c95cb27e83762f7e9fbce79d1b201ae4904184644a4d2703b58133532676",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/score.json": "b8edc17e29455df45f8a808d0b3e93751d6711facb0787d2563f0c54d0b5ea41",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/metrics.json": "0fd9455359a6bfbe29ac4ee28087bdf4ebf7aedf2438a9dc9dbd0cba08668bb6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/facts.json": "208df670cf6f1bd9948823bd88bec5c38a036b3e72870715b49a25dc9383fa3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T08-15-06-264Z-nql9/codex-inline-judge.json": "c7128346a191deef43d5ad0e2989b6b1643cdf6f326dbdad6fd7ea3b7b5a8902",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/result.json": "3308739b7d86d4a13892d159ecfda0e1efa9d2721431b3915c767305e556e830",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/score.json": "c528633acb5e47f87881170d89fabb52b5f31097d2ba4e2a138a600d85743b6e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/metrics.json": "b1103de2369b00c43f51012c74a3d0162db9f349992849a3c3b23c6976350f71",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/facts.json": "83ebadbf4b6c48bdc6b2c1c32cd9c10d0c3d0e393d01b9b88fcef773b3879d11",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T08-39-44-825Z-f5ny/codex-inline-judge.json": "062fabf2324e8fc25989f9b3d6ea3053e6ee8a76d1f78615ac512cd339944369",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/result.json": "0a14e983a465d495e4d7ffe23024c3ce52e1bd3db3e79d0d89bca67b3794941b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/score.json": "2ac9df40ab3e7e6eb030d69dedf49e66de6520a93d3de1075bda34351cedda6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/metrics.json": "ccfe551a3641b5ca6e2eb54700f16cd9cae8aa3fd3a03ee46865b81c79ddbaf2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/facts.json": "cea052e48a87c73460180fc643bc87b968007ec505e61761aea2b9e0b89f26d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T09-36-06-746Z-yyba/codex-inline-judge.json": "0e5e65587d0039ae88ed58569872fbd86fc27579ae5271feeb0f963e9a1c6eb0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/result.json": "279b8742d14a60e1e3f5979b0843a2eb09222816850eeeb9d9302b4e28fbd013",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/score.json": "a92a3c3925f6898c1262b2c00cb08ea28ad8fcc0c2b7900bcc5d25e4d2594821",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/metrics.json": "581526ffd0f7c24e8c71d4705eaea5e442272332a43dd6e4540e6119761338b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/facts.json": "45be98263ac225e8fc61ef33f0a39a41b13cc3761270e84493589881b67d06c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T10-23-00-275Z-stm6/codex-inline-judge.json": "cf01b0854fb705475f2c5c41cca026c823483809e55fca9241846e135c048a8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/result.json": "6f8136d0b5c60f56978129c7048294895f79841533fde1c40b8ce2a03b222bda",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/score.json": "d737edb5ad995573022f6bca4553557200f94e99f6535a27cb41a1eb535c4eb4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/metrics.json": "f8dcff8085229911bfd70554be485f6f8d907a78606b04d221d294517f9b92d3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/facts.json": "c6ae1caf3f0ddd4b5c3927cf0b553301f02436d8fbdce959c45d1d9513fc6e57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r2-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T11-33-33-174Z-1fb0/codex-inline-judge.json": "8c3fa739dbef12156dd3ff89cd5430755f599dec9594c9ff708f85182cd51157",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/result.json": "050bb843256ffe4a710df0b455e5f9b669fc4292e132797cacdfb2c9fcac33f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/score.json": "ab54ad31c9d3e61fb482fece769fe3c533012a11a8a9337a000588e50a4d9f5b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/metrics.json": "ea1dd0ee2aeca4635dd3cdcd3e783cd37d04fa56d6b9ea55ec57fa1e2ae92b05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/facts.json": "b266754f4a0edf8d1a47b72c1f7495124a1d10b8a748947c689b44c7f40ab981",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q4-2026-10-19T12-07-26-472Z-qjh5/codex-inline-judge.json": "61cf37197e8f7cab2dc2a6093f0aed0f9a0c6943640216bc8548cf08796cedca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/result.json": "3cabdfd9ef47c61e39f5f79cce06d5612e9474b5df589c068d303f8f0cd4d0e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/score.json": "600d1fd14cef81d92217db540a58d98091f80437b12d834cbeabf18cd48aade4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/metrics.json": "e91e588056920b4e88172b5eadb325e2da087e18d798bb078b90e14d34ec224c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/facts.json": "897da66c420f133b580c110112f005ce6d564b3ce7fc037dfad974406eb57e98",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-petshop/petshop/petshop-qwen3-8-27b-q4-2026-10-19T12-10-32-239Z-8z4q/codex-inline-judge.json": "48f3c70042a4f76526ea7834a6ae42138f21adf0ad98dc6956e12e6b67ca9a1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/result.json": "61f29c5ee71a16e4a97a9eb0c73c314cc08f8694e35232ab8bde1a430c02b74e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/score.json": "c9492eb361a696d875e810df71491b2e3d11fde5478980ef5e3241ef224cae57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/metrics.json": "e628bb25fe9ae8959feb982e898a9e2beabcafcf5e56a6f0f7b8aa812b5517b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/facts.json": "3575640da59fa8a4e11cea57c69ccbdc0b648982386b41b4417f0a3920c99190",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q4-2026-10-19T12-16-18-577Z-8xf9/codex-inline-judge.json": "3487811a8e72658745e8132cfa105d854ec6303d7177bb2b8759506dd8b0d42e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/result.json": "45fd1229dd24805e917808e146cefda5dc2510dee2b5d608330b744176a30de4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/score.json": "9a45673ee97b0f13f2c97d82e0d04792df57269d2fd30e2eb4b93de07bbb255f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/metrics.json": "4ca24582060e2324e9f75b522048d9f53d92b03a9906fea78609931aa2f78acb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/facts.json": "0ba7988728c5df969a32818ef2982a382e88a6bd69d3484237b8e1f9c3f7fe79",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q4-2026-10-19T12-21-25-640Z-10oe/codex-inline-judge.json": "0b16949cefc75852b46cc5ac63246695c16f7d1f7d02a963f681dd1a373b9f43",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/result.json": "c73c0877acc29f6cd2092ca320def775d5e4ac8180bdbda194456e0f86fa512a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/score.json": "1e6001b3a95003dcd114c6f34cf6bb4e98a541a57f1dc1bb2e8f218beb3a2b4c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/metrics.json": "e6474963140b1efb44255b4fe07e5bd6746273a5445154af5950842deb28915a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/facts.json": "3ac1e3b880806ea4b7ec6c1d170e3c8e3dba1299ab683a81eca44b2b69b918a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q4-2026-10-19T12-25-48-560Z-dmny/codex-inline-judge.json": "fe42630667da9999667f533c353a4010563d6c178ac13093b602221516042d65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/result.json": "c9cd2866dbb3b0581538bd34424fd56e9c8335566976b00f9f1cba46833d2ce2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/score.json": "2b904bdbc57d8f5d1464fe5029bc9c3167cfe901026ad2a233d8c04b3bbce4be",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/metrics.json": "7d34a1f50a83d02159ce40849c3a9ede6d3c1426461e0bbbeecda1d967446bdd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/facts.json": "6c00adbb7879b840d78e4caa429a35a5c0d874dff107901c31da06dd2c73a32a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q4-2026-10-19T12-26-50-565Z-fa9b/codex-inline-judge.json": "4d9791036316857adfa9aac174e7cc31ccbed64f2ca158ca4a71e8472710b4e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/result.json": "e3b1278cc0bfa2bf325c61014bc31beae9649b9af9d7867a0fdeacd637640a32",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/score.json": "aad76394d636b391bcf9e4766135178de4c8266e120f653709023f4c4203730c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/metrics.json": "e0f9baea606b4ec8848808d13dc0e89c1e8ccb0c681ecdcec4cbc81d55c0ffae",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/facts.json": "8a91c6908e66cd61894f740612dfef03e77fd005dd5155d57b4a7a282888abdb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q4-2026-10-19T12-28-02-591Z-o0mv/codex-inline-judge.json": "b6c85329ff505a094f5562143a13dfb7436bac2071eaaace3ffef4e7da237727",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/result.json": "b5e52c895759b65754e69de8397bf4a511dd3cafc5094fcdcf5c3ce02b2dd9fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/score.json": "da191a0812e56ac26d134b27417502bbb4ac379b2bc91ff0bde80b6d975fcae6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/metrics.json": "ff61221354dca047c2d7a9adcce7171dbebe78d5e3e26a4208e25b34c7432c5c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/facts.json": "0798f8a0f604bcb29e442f46c1dcdaaf78540c83768815056be80db9e950ead2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q4-2026-10-19T12-30-55-156Z-u4cl/codex-inline-judge.json": "f57866a693e64820ce08b405c98b037c6ada9ad2965e3fe3d719ee554988338c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/result.json": "9d5afe8b0bbb1a54e712215be2ccdc0275b3879e0433a2393cd918f7b5d33ceb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/score.json": "4ff337765d070280f30dec3ea49cad71212e0d0a58781d340a516dec132a00a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/metrics.json": "cd0bb3f9f8524717a26cbb85e3256cb6c9af4240f9c0094d79080bbf9ccc2c1d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/facts.json": "32414df2da5b2c46411de10595a571fec320fc1450737e72128d33c722eedecf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q4-2026-10-19T12-46-11-942Z-dnha/codex-inline-judge.json": "141a8aec7228da65ff30ace657582c694b62c724c5ca19050e6e862e5184d870",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/result.json": "bf7a982e9d8fc4fc6e72fec252465c1362c6f498df3ed5da1cb16450f3dc2c46",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/score.json": "cae9acaa90470fd742569bcb40fb557ac9ad62d440aaf63250b4abaa78ef084f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/metrics.json": "6131c873d35ef54aca26642d38b0901f06b563aa107e25e2b545f972e50e1c39",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/facts.json": "8198e96ab5012855ef99292fa0ba7d63040a944a1a4a96d5cfadded2ecafa0fb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T12-48-29-306Z-lyad/codex-inline-judge.json": "8fcfc84a35497d82b568a39c355075f5dab7acae689a1fe7e1cced2a78f9ba0d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/result.json": "6fbb3015403882d7a397d671dce3f277c150b8f00695cb9fe2e8edf86182c9ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/score.json": "3d3284217f983c71940c010526f6e56bd94680ccba0f53b72a045d095dee783b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/metrics.json": "bcc81a567c7b2177dcd5a5c3ec6efc23dec3208a24d6d716dfdba0a64ae36b76",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/facts.json": "af7b41401a03c82d2529da2c60960cabfe93f76939c3eb36274508b440481ef9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T12-50-21-596Z-t4h8/codex-inline-judge.json": "cf38845d9d4569ac84763d532a4308168457d24b9b5153a4425e16f133204334",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/result.json": "41901567a381761085608f74c0d01d1c250c2909d54e73c545b9b014450f51c5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/score.json": "c438ef7c7338bd9bddf77c25d11c45ce3f952998d24c3481a632c1e326e1aa0c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/metrics.json": "a53f07901fb00b3a682b968bb0876d20c09fecf7f9acbe13ba8066640829d1c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/facts.json": "840023bc45b7f0d9134a323a20101f5d454652fad5d621f58f8bb9a5e34b32a3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-qwen3-8-27b-q4-2026-10-19T13-00-45-988Z-41df/codex-inline-judge.json": "5b2f8d6d5ba990e5e5474d5e3bb38f9f0c8296446efa9aeff9a6280c3622fb75",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/result.json": "94ec4e0912e59d6e979827a91615ef2233d7d54bd372abe2a1cd025068a6102e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/score.json": "41acf4eb9b4617795a5422ceb75970eebc69b37ecd31eb2bd069c7824d2f0797",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/metrics.json": "1eb41ffb144820dbc95f14cad008228ad09c02726c424636af25e748997d535c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/facts.json": "f1506a087fa2c410fd6662df41aff08c5af26c53563c42e2011982e1b3d0ac50",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-records-intake/records-intake/records-intake-qwen3-8-27b-q4-2026-10-19T13-01-58-057Z-k4em/codex-inline-judge.json": "4e3624baec1af0bb7c47f0c64b744104c7d44b2cb501fb01a14be231dda22aaa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/result.json": "73cc9886ef719069b2a93a84ee7913496584b52ca85f85b353715ab77b7849d6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/score.json": "1aafca05e6890169ef7cd1be7b8bc708e10c00bacfe01558a82b48d8c7328100",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/metrics.json": "3b407154c281f470c027ac2bf5d4b518a99e04aa41284ec168b7d2b05f56d02d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/facts.json": "d45831a593fa6b15bfa09690dcc2e0c8e7f78a01700a7388b3b146f858780950",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q4-2026-10-19T13-13-45-318Z-qz2z/codex-inline-judge.json": "7b30007205f1ce5796449a79753c6e24b3436e39c969de483e8840224b373ddd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/result.json": "bb2c35be2fcd0f4966eec6ae6728f85ab65314ae83ad5c08124c0c97a12cbea1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/score.json": "cb59720d24fffe8545716d9a1c1e280dbd9bbaf60b796956c0bc5b3123060608",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/metrics.json": "e6432addbd576a64b05c6e0da9284e43c2a1f6960a110fb669cd72f53937b7e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/facts.json": "617f199def534453cc4a3b4d5520ebda47881dee7fec7a5e5e744b488e1a8f90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-qwen3-8-27b-q4-2026-10-19T13-16-22-825Z-jh8e/codex-inline-judge.json": "4ec195e991d4fa63ad28ec552d412ed13add66d29842026c1c7d879367f31fe2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/result.json": "13b4a3543dad852abccaef1bd1e667ece349a56b6981f6a7ac01cd10215b6adb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/score.json": "639138fba89281ae9468acc77830f028a0ab5d905d36e605117b603e8ab9d370",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/metrics.json": "4de35021d307912937f33db4ad6e54439a4fa4817ad8f412c34c0a432556fe2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/facts.json": "0a6c3fe6151a876fd20826a83bde700fa5c04e7edb3f527ba9b8a305931d011d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-qwen3-8-27b-q4-2026-10-19T13-19-10-698Z-mtf9/codex-inline-judge.json": "5b20eea4d2e46f5fcc1850b6166802f6bff67d5c8e629384afe5d4af20e3d0e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/result.json": "caa29f200e1e7d39575ca713c1448e43e822056a3cf69fa7cc1519e896078a3d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/score.json": "267d128fdd8eb52a9ce9934044b5d5cebe46db26044ec6adcc62cc6f2030f73c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/metrics.json": "a307e3424f31c0c1c748fd6828ec577e28ef1ae51b0c55b497f56aee02b272bf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/facts.json": "380db54ee67ce909a244f6d69f3d5b645b6c0984221e3302d1e5edca3c9e420f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q4-2026-10-19T13-30-07-253Z-6hrh/codex-inline-judge.json": "968cf702df19dee02ade9f5371df7feee5c2e3a7829242b2e0b7cbbe27a6ba2f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/result.json": "19ddd8c3bd2ae9045504c324ccb2d0e10103da67a0c9196b13eb833fd2e3396f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/score.json": "640d3cb07f971b1a6e1b5ef4bd7d455e8bbf841901e1d4c46568c67d3e910d7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/metrics.json": "dded6e1dc7e41526fc27e7d878e975401abc060b61dfae551386cf753b59e8b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/facts.json": "089507150ac91bc7616e190b41a559428ba43e8e9cbc176656a5b24b14f3422f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-qwen3-8-27b-q4-2026-10-19T13-36-30-696Z-6gqj/codex-inline-judge.json": "48e9758698493715c8db7129537aa5fcff8b35d11c8440c62c34b8cfe70cd754",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/result.json": "84b6d91cf88ae1737316a8827d1a9f836199f1db2a683d5038dd43d30f17b63c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/score.json": "b45089aff64c03f91d1099089789f1181b7b1ce9484377aee143c6137b6a383f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/metrics.json": "95e2b73a30b228885362f7bc8ef965235eb526e514bc1d10a750b2cbdf1f472a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/facts.json": "aeeda0c7eaab73cfa724c8708c4ffb5bfc5ca413838174446bce3083f31881f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-qwen3-8-27b-q4-2026-10-19T13-39-00-371Z-puw6/codex-inline-judge.json": "5ad6a3ac8cdf898f1fa7cd9c069e6f10930fb3270968c870d47610dcce677f4c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/result.json": "e353915ff776e609881c10a53634840ecace76355339ab69f523aa805cb7cf2d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/score.json": "16ce7867738eaf5773578aba4b6084aa17864ccceb9f02b40f743cf33bfad82c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/metrics.json": "457ad9c3e1a034c205213dc25bc52c56ac11acf9e522da0138205735697c8873",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/facts.json": "a7acad716652d9e7088803d86bf01fdc618b0e81f4253c8578cac3c99286e6e4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-qwen3-8-27b-q4-2026-10-19T13-48-48-025Z-k0xe/codex-inline-judge.json": "5970ae4c9cd800174db861e75fecc5338720fe37ce4ce6d453a58b6310dc3654",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/result.json": "3959f117145561936cdf9d58091160dedb1f974bb73cde102fffedf6e230acf1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/score.json": "5914d96f6c23dbce4b06ac101cfa69d9786c6863a73d8e67fcb8707bb477f1ec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/metrics.json": "3100e0ea6ae729d7e0f001dc880d70b0ac80a1ea9e3f47ab2d47d9a4c5d980b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/facts.json": "a10d11e7c56ec4e4cf1797982d81d43c0a952a8c624626f587257967c50c6377",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-qwen3-8-27b-q4-2026-10-19T14-01-03-377Z-g7e0/codex-inline-judge.json": "4a6d63c5db247edfef37ea0118880c033dfab9047d6807db93057c3873f9dae0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/result.json": "76f082538f6181afb4bfffc2fa5fce0e3ba5f4fe09d288ced6f78f185631890d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/score.json": "f637d4312cce2fbc1a54bef5b0909f86afad44aada66953ac2be4e56f88f6661",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/metrics.json": "7fd24b274532d0a560068815bf0940781a608ff68e70832caec9031442432760",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/facts.json": "2624943638a94eb5c2e6eecb37b532e77152f89a85c7df31d693a26de2cc74a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-qwen3-8-27b-q4-2026-10-19T14-32-25-885Z-eg0x/codex-inline-judge.json": "355aee41d801f9fe0404feafc40bb5991ef3cdcbc70968c2526037c31129aa36",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/result.json": "39c108a50b7b90059fbe9ac220a2bbf1a3bcf173e6c73f9914c42f7eb1f472dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/score.json": "a56fc96c25e97416b3f0268c8635bbddea4e06c001febb22d92a047459ed989c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/metrics.json": "7f364eae7ddff18ee5cbb45c1c350392bd6169c1f4412f8fdaf8d4d2efa50a7b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/facts.json": "69414f51f0c5fc46f16cf0c5ca976b5d24307196de2f8fe9b888fae0db3ced88",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-qwen3-8-27b-q4-2026-10-19T15-03-37-950Z-xjql/codex-inline-judge.json": "7d6a29395c7f8adbc91ff09f28ba4a079b6df38268ee6d7651d7109bac302d06",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/result.json": "8e7472d9000e79603d9168befc1945b6cb3f6e47b69e43a094faaa0b8c4c6695",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/score.json": "f1cefa63f5b84fbb77292074310589e8b867d766664294f55c025345fb3f6bc2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/metrics.json": "a0b81825746aa7683d65752e1437697c1bd0e97640e191f5a5b8f7de011a41f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/facts.json": "fd7b2ea271a388b103f171c2a5cc5a8e1a988a5988b53f70a774d513f9df850e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-qwen3-8-27b-q4-2026-10-19T15-17-39-426Z-tb2j/codex-inline-judge.json": "3ed837787dceafac1ad9e2ce5371c9d84241cbaea5d8f732c17f549043f9f58c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/result.json": "90ab34ff31f6a8e3805dadee2f57823ae039bc70bdacfe145e32adf676a671d6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/score.json": "685e4cfa88d0c517d2e3a6e97721943896b695d2da6c6f61c11e22c47f4856f4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/metrics.json": "7ae6e36a07f9c0df01ea7c368f5679521bd4a93e332b9db1c4a2bd1f16d9084b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/facts.json": "154003bf18b250a65fedede4bdbda7246ff04e4c8136d6fa53b8baf445ff6a10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-qwen3-8-27b-q4-2026-10-19T15-26-06-425Z-d839/codex-inline-judge.json": "a50479493b1d823d6506dc8681a55835f290d95557e7150a254f49c35db13487",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/result.json": "403ea435a4293a47fc60fe005e56b4e2eefb73bcc968f249556a6173045a4d80",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/score.json": "f04ea47f95bd79d4ac4daf775c9c31d38fb954a8c815611825e39f94033216c5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/metrics.json": "c2e757e67e347a722204f14784dcd32c798603e94369605bf624f2072bfd7fde",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/facts.json": "b0eb4d3da8fc596f1a0ef8166d4373aa6ecb1b93fc556e2837ce7cf712525687",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-qwen3-8-27b-q4-2026-10-19T15-29-13-943Z-jwz2/codex-inline-judge.json": "7371ae19faafc252c01174e9273a3f3bca4d6aa5d46b899df3a0803865cc6c69",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/result.json": "66159765356c238e75de990f8690e551cd4eacf73f32b7d4954b7f29e76e3a04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/score.json": "746e131ef2f906e1b60553789bb331c9e435112791d2d61e22fb685aeaec5099",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/metrics.json": "92faf1d4d61dd68ce430ab370d22ed8c2c1ff890a6f1e7c85463709ba463e568",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/facts.json": "0bd02f26d5933f1c09100a5a7a5fc6a1f6eca9c6bcffda2d557951914f6be94f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-interface-contract/interface-contract/interface-contract-qwen3-8-27b-q4-2026-10-19T15-30-41-095Z-a175/codex-inline-judge.json": "63340462eaf1fb1fca8838d9c812a20ee8cd62b6d3f1529e10fafdf36756dd6a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/result.json": "6f6b479b1f99a27a17315e25a1b9aff8618aa7a061d43478c7f415cfc9c9c73b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/score.json": "8da69996f8964e88faa5828ce6778eca765f4bd1fa3e15c1f340f6fdb59eff81",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/metrics.json": "3004c600d0c8402ca4c3c9e64bb8fb7a9c616ba389df7697e42304ca4eae2ecf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/facts.json": "86573c300a303caeca367e0168dd04f29cc9f9e29e89c3f89f4fc2cdfa7adc40",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-qwen3-8-27b-q4-2026-10-19T15-35-31-027Z-64ge/codex-inline-judge.json": "1c2f243e70c90c166f6d3836e07300aeb43a8584d183ca770e5b35e1a15177e3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/result.json": "48a0810b41c8dfef6c904b1bada8b6fd982a4ce436a688dabfdf588d9b63ac2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/score.json": "b2958963faf111ee5b508d931939a3fc16708dbadd6496c7f899b2adc752e641",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/metrics.json": "3307d125ac6444766034ff8affba98ac9a954ebb241556cfa2f41780c88a81b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/facts.json": "ce39ae410209a0aabafa0724139734546fbce8440ce52cc39af671e1745b2b2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-qwen3-8-27b-q4-2026-10-19T15-42-13-494Z-fepj/codex-inline-judge.json": "4add1f918897a5d5a1de602d54bf26d3ef47f9823252e66dcb7883aed9f9ffd4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/result.json": "50a01edf19738512d731547199745ef7321dde541f5298c7bdedd6678422b982",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/score.json": "4ad032a0b9c239ce774017e5381ea9cb087bb62ce69b6347d51820163fbb1d1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/metrics.json": "5824d283b5e0598e286a119aa217f91e300a242ec8dfe894235848a609a67923",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/facts.json": "bf622d3c90dbe99e47219ff3c49df571dcfaebfd979b03ff1d36d9f89a8989a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-qwen3-8-27b-q4-2026-10-19T16-15-19-946Z-guu1/codex-inline-judge.json": "c550ef85fd091e66d5007757b6c371959eb0b502d1a6dbd795a9800712fff69d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/result.json": "5bc30a3c4a7ecfae1b13c8bfb690dc214b9017bf8eb455887a7aec6932bbf69b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/score.json": "e81413e958f81a5bed5a217f24e6cc39c390de45c3d6140f9a6968823d3f63ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/metrics.json": "b556f9cf98d51c833e5a1ce97776b60d0a8d21e8008a62075ee2b449d5628d84",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/facts.json": "eba782be6978df3777a86f7480cc42d0e198dd72fff78dff962ac3d5d404d8ce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-large-pr-review/large-pr-review/large-pr-review-qwen3-8-27b-q4-2026-10-19T16-21-36-254Z-0172/codex-inline-judge.json": "082cd0e1b7e06fba480ab97910fff22c03b44ad10ef465e363b1ca366853d7b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/result.json": "e747e8db24926e1191105a056b62ab256a268eb31d07f7acc96c626cd8de4408",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/score.json": "66a816accf3c454fdc8af7fa9df49ee31303e25d34310ab6c8f893ede8d83a3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/metrics.json": "6e5fe405287a9ece2e3a579c091a6122d2bbe6e68a3639ae0b896a67427a7939",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/facts.json": "8d080f6bd645208d11e9e1a651409c333be051992b5ccba558f4a6d63577947c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-qwen3-8-27b-q4-2026-10-19T16-59-01-589Z-eqxk/codex-inline-judge.json": "4620de03635e2066dfe7c47b0da0624bda6926926b3b3c5c3f730c70b14bf960",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/result.json": "babde97f911d2d462e6650a4b0831141f01833440eb9c9b01f9f42759b0c9701",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/score.json": "31823ee2a1b1920d45c1dbc9450bb2c7f4ff0ca70cdcd9f18079716cfccd8ca9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/metrics.json": "1d4856f26d1e10ad8ba55ad0c98b7eacb78089c1abd7e7d1c0abdab932edf309",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/facts.json": "b3641718d900b5841897c47d36f77e3b5f5d5821851564d23b26b33d83e9a122",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-qwen3-8-27b-q4-2026-10-19T17-09-35-000Z-m6d9/codex-inline-judge.json": "6ed47479df950cd3443065fa2d4e1220ad6ad959347be6a062dc053bb44a1b0a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/result.json": "88b979a4539f5d7c40db7405653a062f074da69e5ad72c86e8c2768c3ccfcdc0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/score.json": "b389624b4904757660494a21cbed2276fdaea626d518ce732a1543df337b71cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/metrics.json": "1559acefce1abf09db08964718774128a9cab7bc0b14166971aa4b45e243f34e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/facts.json": "1800536fa0dfe269e9ae6d0023dcd7b270f0b71a4f72bba9463bde4774cd9313",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-qwen3-8-27b-q4-2026-10-19T17-44-21-871Z-wbrc/codex-inline-judge.json": "dfd2ad85e7c34c29eccbbc4196de215a9efbc209b5fbd1c302e12b10c0ad98a3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/result.json": "ec15f68c9b823b412c91011fc1cec921f2a243c5264e826d5bd6be6751851065",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/score.json": "4c89aba61e9f8e4c7ccdbcce1c8a679d6e486bc018058926852e008e97d13d96",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/metrics.json": "04bd924bf40ec06c458acc662f6a97a41967d38e71d2a1ac34dd837beddd43fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/facts.json": "3228612482f4b12ed2d8641a0525fbffd2ead07e7a44fb964d038f410872e2f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-qwen3-8-27b-q4-2026-10-19T17-58-07-087Z-by6b/codex-inline-judge.json": "3c8ce53d02a39a392886578d18a14dd86efff3ef41e74568fd5ed1ae054f5c57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/result.json": "e448f6844bf4195131f0c20028beae02805ef1747fd204dd277e7a6c3fbc329b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/score.json": "6b52734124a06d7f3c16fdd7f939e44baf72a37ede60fe1300712f2c1042ec9c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/metrics.json": "6e5a7f569eae0a359824bfc8bf3744845702bae2af6050ab5ef3bc6ae202736c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/facts.json": "60610f959fed3160185feb40b1690a0f305e5917a563c7bf6ee079d3b9885232",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-qwen3-8-27b-q4-2026-10-19T17-59-27-100Z-tmi1/codex-inline-judge.json": "8597b0b6dcabe9ca12277294f93855215bbe98d94df99785a7bc9467468f6338",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/result.json": "be19e362f238c8403854eb323d1863e1c022c5f59ea52da206f21b7799d547bd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/score.json": "daf0c1fe0c84149f95e0ba80d03b13ad23638fa1ee104b439052c9475a43f7a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/metrics.json": "5525a77d51fa37cc8e6f944e54ab52c02f3ef98f2a7d55e1bb3cbbc23113c77a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/facts.json": "3c633f3d2576a6a3610463bfdee1af3d7fc19f0082efff14ef96e5c13f708bf1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-qwen3-8-27b-q4-2026-10-19T19-01-19-897Z-9k3s/codex-inline-judge.json": "ed62469f973d7e0aa39ce2fe233abbec073a51cc6cb72785d5f6506d453d1a13",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/result.json": "f6c1a607e6ddd69fcad68334c608d1e56af49ced8263b9028f631c02a52ad835",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/score.json": "1c61d10be8766ba8a445539593efddf49ea461d470295ea433d3df5e6c2ce481",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/metrics.json": "ff81c311727ede16a90f16a1b54d7ee70ccfe05d11a7d6d72972b046e3bfdd1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/facts.json": "3255e55cd91ec254181d86835bfb6e5a43b3ee9f93e7455391ef30aedbfafa83",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-qwen3-8-27b-q4-2026-10-19T19-14-27-073Z-ro6z/codex-inline-judge.json": "955a4c6a7b4c74df19090be17324bf6c8ea7255fb537dea6b0a9866c5b605f5b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/result.json": "cbd8a9c40f9cb8d5f66b230bb7a30dc56b10b68a8c65c85eb491b2ed3a2877d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/score.json": "2e1b6d386c2ea5dfe010df3ffadf20d4b66549e59bfcf82ea116beed9d8d9660",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/metrics.json": "3d41dab231d2d1776d7b1cd2fc775f626cc5d68e97aef9b2cad9f45f9d2d3161",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/facts.json": "d37e3776234728b9a76fb20f963c1db83434ae53ce7780ccb2414cedab203510",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-qwen3-8-27b-q4-2026-10-19T21-05-02-403Z-vzyo/codex-inline-judge.json": "863cee01d02792732e3d0fba77bf3e9a1f359faa99b5cfa82bb517c015c9557e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/result.json": "dffe3bc863035f7a6463ba3a84daf22079b4607ba2147b0ccb04db83305f6038",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/score.json": "a2023b84e3ddb6dbbee48336d76800de936767b5b25a948a1ee54fc283a1d055",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/metrics.json": "68e721d350cba4ca442836cfb756139cd96322e854b171d5d56051ae4a4f6507",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/facts.json": "f6cf270a99b4da0d332e4738955a34cc3509db5f5ce9b2bd2d6e789c21081109",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-qwen3-8-27b-q4-2026-10-19T21-50-03-988Z-10tm/codex-inline-judge.json": "6f8a31a33d62bffc315e3299b80c1158c5a59d627ef432a75c62db1f57f6ed8a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/result.json": "0e8f4cf33a5c8ba8b6ad9ca2ef46e4f3855b0461c3dea234be459c3f3b2c3b33",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/score.json": "817caf5a89ec9ed586c9a626ebe0c6fbefc891f86694c94dd571e87462e69091",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/metrics.json": "164ea52d182ae27b336f739018719be7d9653885a5bb592c27c40d05ff8d56e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/facts.json": "38ed9dfb70648308548f80af17ff84bfdc3e0f1f89514d2d50534e70d56c1a72",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-qwen3-8-27b-q4-2026-10-19T22-32-56-183Z-tgru/codex-inline-judge.json": "662fb726b947a3a985a6a6ae71bb25a70f1507ea94e310a0c52bb3f0c38335cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/result.json": "24be607e32c229be2527925f697ec12c15763e7633c61a87a17c9bca1be54085",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/score.json": "21e5cb21ed01d30158c1b6395ce630ca04cff700b901c4b54f9530376afb51f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/metrics.json": "9690f69dc07c83b5a37e4d688935bf81018b82ec6fdd9f2be3fa8b7b6199f95f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/facts.json": "a023f96a051619e25aac85edf6e5148ff5397e77d489d54efc723fdef1a9ec5a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-qwen3-8-27b-q4-2026-10-19T23-29-54-614Z-x14f/codex-inline-judge.json": "f24e39975f369661bd8837eaaf9d881f7af56859d29f443be8a69daaf2c53cb6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/result.json": "6777782e522fc3efe49ffd0ca65163ddbda117cc96adc5b3403946f06cec9d2a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/score.json": "7c3f89c4e831436f4f96b4aad5bfb45892d4ed143bb7a45199d57743e8c206c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/metrics.json": "233d6ea290c471a0e96789652e5536112f3d99078fb2a87552639f63fa2a682d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/facts.json": "7526bddf814ced71a490862bd99bdaa2ce66fa9f650e8ff5bf61ba2f402f6eff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/qwen3.8-27b-q4/r3-qwen3.8-27b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-qwen3-8-27b-q4-2026-10-19T23-51-24-060Z-n5v7/codex-inline-judge.json": "8b13ab5d6c757e977d8ec3d0a92e3008d4d4001bf39688b8d7bf1c1c61e69f45",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/result.json": "4342bbcfadf5289bfd0fcdb447c8c9940a3929553dedb43ecd06b8b8d9c62c5e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/score.json": "a23dd9f4a40d9249cc2012d31e0a05cb55b7fbc3a53d80d278eecfaf9c2cf766",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/metrics.json": "ce11895ee56003bd0293ef691b1364179b17ec3b2e9ecc51202c98ee00540838",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/facts.json": "fcbeb59d1ee7c689725b1a2696dc4c35d11be54e539d8ffe4f3b7efd548a96ce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T00-52-55-513Z-6atm/codex-inline-judge.json": "900c4f2288d5960ad784bc05fa2e730f545f0dc61cb63118a597e5b94a6fd747",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/result.json": "1dc13a6a58fc644ca216f1dbf3ba8d9bd204340e459a2328038fd044b4531829",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/score.json": "f9956dfb89d054ab46980172e366f5e957e436ab3709141b841408e8a541d444",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/metrics.json": "cd077018d716b5ca98483067df085af4394f56cff3e0be428a398506ad56e0a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/facts.json": "7c5b934977e7f8ca682068267f49b12036e745d0bde7ac783a70d36d51bd945d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T00-57-59-049Z-siiv/codex-inline-judge.json": "08ff8f154f12ed7afb2d318b431af0bd008ef9e5e42cc593d9573614215ba04a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/result.json": "becfc002a8373c36234b45b4f3ac0a3a76f6ad1c194062d5e2d90fda87d486a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/score.json": "81f5737b0f18cda01ed1538814602d5d67388926ca382e11ebe1bfe8d636a1fa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/metrics.json": "37e7c4306f23b4d21b61c59909295accd89975675afe9bf2df00805be0e87d04",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/facts.json": "f14bc1dc41c8c00c67ec00275bbe51feed939c87c6f7eed6bd1d1072203f5dbb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T01-03-27-386Z-x1mm/codex-inline-judge.json": "412694c76c89bbb1f2f74e4fb4a327eae39c26c077f854b8619b25dc178fbd7b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/result.json": "0a84e1e94c6039f2bdc899e12e4bb4161f28fe59611615020351cf87b59d8985",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/score.json": "47358b8f4138b69ce3ff4bc8b81b8bd3df83fc68a324969a2460b5b28baf637b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/metrics.json": "4ff8ef2dc654d3b4dcd99baa27330a5dae74ee520bfad9675510f553ea2b2f7a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/facts.json": "581c1eca6374ce0279e9bb26080f8ed62f045aa1afada7efc571b8d169768a6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T01-09-32-150Z-3teh/codex-inline-judge.json": "1c7a000d4bae492ecd261c1391a48c9c9ae0c4a39a9ca6f7afc2df000ac76ea4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/result.json": "6d68edbdd13f4dc73679397e14efaadb28d6e7887815db837de2ee9380322f40",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/score.json": "1ec85f0d56a43c5339c39aecf0f2c2dfe4d44ce289e02128ef31f76d5fdefc0e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/metrics.json": "f2c401f007dea3ef5bd96068dfcfd153ed63c3e1f838471dd7f626155532b487",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/facts.json": "66777e96f76f208cb0433b31994a1e7aa4dfe0a4a175508cd9a6fc29b8fbcd44",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T01-14-52-857Z-cq5o/codex-inline-judge.json": "d363d064c3f99ad19930c2866452a60fcc1ec166ff003b8eee97b4d9b2738a07",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/result.json": "a9630bb88a6e8521def88c122e0d13dc12b0984438c043488ccfebe4c283a6df",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/score.json": "24ad743e9bc4540b612d2a271f05d206e482967b8a2f99644ec94f2886730375",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/metrics.json": "2e9dcf11445ebc1aebde6f0865f69b1727f2a5951845d6c304237b9926168472",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/facts.json": "8fc978dba14967dfb0d4645d6b419a51230f21e0a44ba39cce44555d963e631c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T01-23-24-971Z-lmwo/codex-inline-judge.json": "3c144c8130e3ad42bded5e5d3ed29fac5b90afa6e99a8bd47c037f74e15cf246",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/result.json": "eb78bd67f62b5a5514090364c9b211f2f98c1d66c713d040cd53e37971db0ce0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/score.json": "89c3928f8931831e7c958eca25caf78a3a30be3cf0f1782a0d1400b9bc7585c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/metrics.json": "bb7f2103abf9f47f3cb22e9fa26759a2bd72657a6fd4c61a77b644f334ebb39e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/facts.json": "5dc01856bb37bb75798fd2ab461c4bbe4fb8e48fb2dfd8cd1a1a169877b217b2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T01-27-58-250Z-t41g/codex-inline-judge.json": "0992a0241e90ba73ded956baa782607f96280626192a507bd38d1ae8b7cc92ee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/result.json": "316a7b58c4b257ccf2db3ddca7f7716b2d3de8377ccc07d8b757aa35356506f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/score.json": "178fede95228361749861e7daf0ce5ef5f42451387ac6f0189cef0f30012d290",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/metrics.json": "fa0037c5c8c4f134123c281f138b23ff96b67f7828f37be9fdbad72bfbf28209",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/facts.json": "4f8b3079111ab2f94fa49e744f9aa6a28bbb5cd973cc55b7f0ee7357fe581795",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T01-33-29-511Z-sh8k/codex-inline-judge.json": "82e2c35e0180819a1717791c5cc3780404c21aa3a51487a2bbf59bf4b945a551",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/result.json": "a7284dd8444a2eb8e21c7ea24b34c7fd3695535618f2799b70ef6a9dcbeabf5f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/score.json": "634e5f608c8012b22cf0b38d7bf4080c175a9a56a8ebd220e14227f3cea1cedc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/metrics.json": "cd05d6568e1efc93cc911891082030c92cc8d6957b3977f017fad304ab737e74",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/facts.json": "1561fba0ced6cebce2f9f60ffd067217f0638ca66a9db9aa70206452980fbae6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T01-39-15-329Z-wwav/codex-inline-judge.json": "0b0a7ec3f769c23c241ab97c116716d73a79d805b1dbbe3c8a4c1adef53ed142",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/result.json": "5a06739a2ff573b60d275866077d63b0ab9248e0697b0f50c13cb41bdd4efeeb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/score.json": "0ab9b69e17c68af032c0b27d9d395589c8e0f64367579715d7f4540cbe6ea988",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/metrics.json": "059bcbcf956daf40bc0b1854c03d9e77e3ebfeeb98eb28c6788e1e6acd165de3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/facts.json": "49dbfec02191579aeca5db0846cae2917d6c366a1dacd79a437990486e3748b2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T01-43-50-892Z-scgx/codex-inline-judge.json": "f95f42eac340293c83ad8df018927d4ce84e4db0d9917d54f103da81e74df143",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/result.json": "c4506b9c875dc975f4278a9769128701900fa9e927b33aa6879a3f0c6276f20a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/score.json": "5dd691cea4b8fda3a741e27a9d18ae3eb85240a63683719736fb38afcca6c4f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/metrics.json": "3f0061078c65289663add215b790bc77270a14c9489a2b0eb6fd06de3d402dec",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/facts.json": "96d27cead59de5cc2d404e4d74069dbd2062ee918f007e371ebfd328b3a7a6e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T01-48-41-400Z-2slk/codex-inline-judge.json": "13bd270b18a69760533a0b662d2ef7810c2220e878818e60640607960b42e165",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/result.json": "ef6d92ed11a168b9fcff3bbbed9dee2509ec85938c7b9b7f3a1481d4be7694e8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/score.json": "92e7df2ee7d83fbdf3fabbca928536a38aeed07d0fd05a69c7a0badb30fa5d2e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/metrics.json": "c4038dbb5c78614ef5ce8ba369e12104c51357889191d303d8aa6ae24d9b56f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/facts.json": "090e683184dda488efb11ebf3bc53bf0b4aaec403a9145a21990a2379ce2234e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T01-55-12-670Z-pgew/codex-inline-judge.json": "e499d23bd0ddade418f2763c83cf4edd5e904080cc169e6af04041399125e63e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/result.json": "3a68608f80a77455489b4cdd2978789deb2f7a60dd3b56f6146e4cf9d002da28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/score.json": "34c202e85a48c50790168623a950db00150dedc19489ab3773b8429fd3d45781",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/metrics.json": "912726cf345dcdbd42aaad948c36109e29005adca1d26b4be17c3baa8cf5dcf0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/facts.json": "42b8c58cc140d4c41fcb363db5a76f4238e9091d85d356c3892b11d85743f801",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T01-59-40-749Z-5aiy/codex-inline-judge.json": "827c6f3de0ebe3130325426bbe27168845049de837de97fb771c8990016fcce2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/result.json": "6be765323df893a905a0f67f1dc0d24d42d0cba02f4002f621bc0e7cf33c4413",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/score.json": "02fdcd392c812733d6f727e15db718891fe94bd20d7cf2c28892d9200f5e1208",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/metrics.json": "bb0fe14669244c82bcb2dc7f690f0df5844639f35b7f74300944d5b091a2ee44",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/facts.json": "29470dd533206d60a7411130013e2163c50cfcee9a4c45dca43624bcf84ba5d9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T02-05-24-210Z-383r/codex-inline-judge.json": "e1141b8b32d6a68cb289a917d9e572c0f86c54ea2f608b7c61ea69feb65dc4e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/result.json": "be38be2b1918d2c62ca56143f1c45b2914122ba0ce76d5597628662a6bd43c19",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/score.json": "da9ae2907789a02484585b54058d9ea4ef1e622bfb66baacf5a4390e9771e006",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/metrics.json": "e480a4a156a5e97b66b9277d8cebddb14428d71f010650d8c69f01eec44fc75f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/facts.json": "aca16b174ed70d6766d5ce059e06f9e3e8d3b2caa50b574bb0ffabc919ed875e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T02-09-48-993Z-gesq/codex-inline-judge.json": "5a70525e5131a3fab0192aaae91a601166586260d0d1e4cc2fdd4bc2c6ad2578",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/result.json": "3287389e88c0143bbc9e33d0425ebc9017ef76d372bbfbbacb60a469d3f5a28c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/score.json": "3af7cbb0a8a7350bc0b840f06a547d692297afe3c0117c665b225127f3e2a0ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/metrics.json": "ee59ddb420e20620d16d3eefad54b1dd957ec0ee5bd99f7862d5eaf226a37e99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/facts.json": "488a0d8ef4a9d25037dec19357528e4667fe19b3807010408f25cbc54fb1d771",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T02-15-48-886Z-030k/codex-inline-judge.json": "8aec88fe5763537de6cdeae212517f8e40175b82e4256cf7e727501e752a00d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/result.json": "370402912e8d2566ba8c909b712f97060ed70d791edaa000aec79861df041853",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/score.json": "ee4dc447ba0fe52002cdc27c9bad2947083b7060ea8a104b74ec53a0d14743cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/metrics.json": "c537eed5302a8e6a9e42e7f1b8cfc350f1c376ac1868577b39652e28e4e076df",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/facts.json": "9697de9983c62551ded72cd35b47419f737fa2ac3ae8ef3887662447ab452914",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T02-25-49-044Z-ybho/codex-inline-judge.json": "1116e2106a96c02c35e5ceb8206e256a060b21ccab7e31a669b016ad34e1ddc0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/result.json": "2280401ef4fb3431a62ce526756760b03168a9e5f611dd21162cc87d56bff9f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/score.json": "de74a5f771eb989bf692dae7c1d1af3864b966fa0219d8f4ed87a2739856923d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/metrics.json": "934f859882806cc1684c46001c712e4d47a1889ac0e62285d799b76644183c32",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/facts.json": "e2ee8eaa7ad3d6f97a6e5ca16a46a79ddb8d7cfe140eec9499b35e533b8a9021",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T02-32-18-424Z-oiro/codex-inline-judge.json": "b9bf4b568f4c4608b4b129bdf566a69af05a9fe4717e12eae9222135d5dabb6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/result.json": "0481546387936200979a663c487f5aebda369cc5b091c71f3b3778908a240e57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/score.json": "f272cb7a373ede62a0ae4e382d6f3faacf9715186d7ab15d8e6f4529a24844c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/metrics.json": "057bfac880615105d6e1894a38c3d7e8ed6da7b0ee7d69eda3f4ae886397bdb8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/facts.json": "5e358acadb4c7a3e1ffe4023d173ca92f7c7f10113b703e0a196290ce4011540",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T02-37-09-884Z-f5z3/codex-inline-judge.json": "b36eb45480b65137f52d85361e813dc2c9af05364a3e7decdfc9adaa80a4f448",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/result.json": "dfeb14c9002d838b59a8aa45a159e4c5c355677de1cbf9bcdac629adf8b6fca9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/score.json": "1dd444411cf6a0a28e80cf5a7de27ef32ab56264fa80eb7c424996c561f0ddc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/metrics.json": "03fb53977987591cad203d593dac29557241ed03c5aa0a148c0db2ac4b1c1ec4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/facts.json": "e1ba6d47d8ee03800fa04812f56a7ef1ba58419c11dc8f43ffcc8aa81ff134a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T02-44-06-603Z-r3ex/codex-inline-judge.json": "bc71defceb8c6e237f04f7e8fc70fc01c46e621328d538c534dfc70a2dfe4f76",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/result.json": "520288cdafeca8da248961005230f228ca755455c24892f373219f01574e1c10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/score.json": "90ae2941645c7db262486f506ff8a75e58bdfd1f75458674906e3328687559c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/metrics.json": "38d345dd5e8d262e7b07558014a5dc2b0f2c61eb56602d894e2fe4f1186cefc2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/facts.json": "cfc50d2d0bc022eae9db1faa19c73cb164991982eb754118a9591492c33906bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T02-50-22-960Z-h3n8/codex-inline-judge.json": "6bc9c5e1fcec7c801effce74c983ff7603a315aa587b38944e3491dc586cff3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/result.json": "a863024c740bb727cd29b1841bd9a9f1ccaf2b69551cd8369afec0f8785b316b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/score.json": "3b48ad9ed347360ce20522642e7e57c4f03d456d286827bb49304754d887d7dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/metrics.json": "b3e3abc0b18d3470a41386a2f9f2575d7907583923d004d7d98618958a385135",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/facts.json": "efa92925af53bc0e21f1403274468ad92add178e2a8bf9d26e757187bccb2f65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T03-24-56-389Z-20vn/codex-inline-judge.json": "0a5d027592ef344cf923af7e79cb22c3707b93c838b0fbd626688d297ed61e1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/result.json": "e9dc160e0b126b72c6f13d2aa81872e5a2f074364e3feb9f2857d94f60bc8512",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/score.json": "ac0292f7ba22d014405375a3e535d0de4276b398b1c7d60005ec863d633d0437",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/metrics.json": "bccf7ee195ad719007e41a75936e02dfde0a0030d859707f70d8fe7a9150c7cb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/facts.json": "a97ccbdf822a8d29a72a5e02fbfcdf1d33f6f2395db109a88f933dd000c2b9aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T03-40-51-772Z-6qhj/codex-inline-judge.json": "77061be4fe8a56e0e0eb9e456b86006f5f689446fbfa1689b6dd55bc9a1c2972",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/result.json": "9b7de6e143716ad9534177ff2cb24ed6b388c22fae6889b311bf0a7a627ec66d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/score.json": "6071a11ff4af9f9722039f7ec991970ca672f6dd736de318211a443434f79bbf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/metrics.json": "1459f42619a0a86ec989540d6f6e0194eb6af16f48c4b4ca2ab06204a2adcc67",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/facts.json": "1c28ab8fe24fb60c73609faec61f016b16acc99daaf96af3f61bfc9402839979",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T03-48-17-438Z-cxbk/codex-inline-judge.json": "1c231dd00181afb515718ea7d42357e17e65f168977c32f700b1f6a487a7c1ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/result.json": "53b6e88d5e52252271839cfd44a714f4fd2c96bd698a43b8d3c0b1135afaeacc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/score.json": "bc9a606e5278b295ade11f56b391ef641a928fe12e665edfbe2260118918366e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/metrics.json": "abc8151b29e80f36e8da013b66a087bb6a463c1028066553fa6dfe8744376304",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/facts.json": "454ce86d8e931cda3e70e89cfa234522816ab05f89a55ef1a91733acb5720009",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T03-54-40-084Z-fmyu/codex-inline-judge.json": "b24e8fd231f2186bd781ec0fa72bddef11e874b28e72a2ed9fc69c30a5a185ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/result.json": "a81c915363efe836fae70a901f688f6bdd41c08208bce150eb1717c34e8ef541",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/score.json": "99d742c5bea4ca3622a5a8771face7332529b486151f1ce82dd273be0d028487",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/metrics.json": "328df31a0787601860e88618b65ad5b086889e178f1387d77b290d26d28f66ce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/facts.json": "5d6697c04484754b22360348f787f32faaaf45bed8e34edb2c53f8bc026bdca2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T03-59-26-104Z-mzzi/codex-inline-judge.json": "008565bf00dfe070a08a3d8c4fa29c5887398ef13c28b1ebf7c3162f74ad340e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/result.json": "f4d607efeb8b6f8b12f83f736275fc8a0b0fdea6ee0e5173a0f8ff74e18300a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/score.json": "f806b931b36cc2f54139275434fc6ea9c564ff9ed74b01857902dc8ba5bc29fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/metrics.json": "1726882e653aa3a674b84cac8d18501a4aac70bd98cd5c311ead5c11c4a19c36",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/facts.json": "8ae0842caca13dc447ddcc8e4b71e0acb056afb7dc5aadb97c5248ee51b862a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T04-03-51-003Z-x9kk/codex-inline-judge.json": "02171e3c44696214821b9a2e45e9bfaa00410aad8c83008e4dd09ba121115119",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/result.json": "ea961709e190d3097f0e0dd9fed1f660fb75be9b46f705010c69b030b24e0459",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/score.json": "0313e9b10c84d535e5e04ec2937be590dabb5798196fc6d04ce1bae27299a313",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/metrics.json": "bb07815998b022742b8c4f8bf054b0001150dec53572e75ea4508694b5eec45e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/facts.json": "eb447f0aca6359a1395333c37d75ef9713efe61b58555f1263127ff3259746f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T04-08-32-569Z-58q7/codex-inline-judge.json": "b1baa0f3b3473eb5b9f3b10ea581a6ca8b3e2d0201698a0c599706b583dd010d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/result.json": "c8596b04372bb975044daffbb8e3679000c497d52cc734c50be7f51f1bed7e53",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/score.json": "89741655bae89d722dedee2a9909171c95ce572f38a9945ebd53adbb7e388e98",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/metrics.json": "0a56bd32783e3944168696e9c196b4c74e7db82bf12c7f2667fe14bb42efcfb8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/facts.json": "b559de80d957e331568eb457c68fa330673d5f97f6b9474803e0f142006ce0f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T04-14-21-587Z-9lfs/codex-inline-judge.json": "c3618ded325097a468be112cd6471b614b61553c2a81a9af7fbb86c2348e64d5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/result.json": "11d449e5f14840a8c5aeab5b795f3d5bb8d3e903efdd90fd70ea2fe7a2efde24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/score.json": "02b2099050d721f184caa2d2720f91fa94db5e7a0b45e0baaa473f268cfaef10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/metrics.json": "cdc338263e49d7aff34a36e09044a69b208069c1ea978a0ab7fd29528653c21f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/facts.json": "a4694452e4aaebcf155ce632f356515761f96a082f4c8b7f4f01df2e66b4ab2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T04-26-54-580Z-fula/codex-inline-judge.json": "04899a8c2661c8dcedac3fe1bd286eae031b67aacf7060d55d24215950d66284",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/result.json": "6410632d0e40e5777c988f655fda9f624d2eba017c12aac2a493e8bca3466159",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/score.json": "db3cb4066d8a0ca7f67d74da40bd1c25496c0dc9291314d2e425d3aafa1c6749",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/metrics.json": "414f022c5aef7264985fb07e592f45929a8e9200f028303ffa69b785dcce94ea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/facts.json": "ecd33e8b5caef6eecea04eba386bb00d1564935420b2b36888a8f6ca9efb0d69",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T04-33-07-788Z-joj8/codex-inline-judge.json": "511848709148751c42dbe28ecc850a59ae2005559cee9dd08e98e1a904a158ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/result.json": "9e227555d12a3f4a0bb570e9a41a6188c9cce28c41a7ed176cfae8db2c71c6e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/score.json": "4fc0b5bce55dac1eb84a5d400c530e8ef780a010cf978a5f83da342f5442efa9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/metrics.json": "5884adedf9f0913ef47404c54f1cb803f7739819bdcffe93224772bd9acf03dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/facts.json": "78a5717f3ab00c336d9164d0594dea82417cdf858be48613a3f437e42c4e3910",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T04-53-19-105Z-nawb/codex-inline-judge.json": "8652a26568093646f11678c93db9485a1f4d94395523f7fb1dc891ae8cb8e4d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/result.json": "bd47e5175235c1c709fa090f420135a0a707a90939b4a74605baf656c64eaca0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/score.json": "3b98a9bf9f49af4f2cc76ab12bf8b1cd53a05ccec7916d78724ef4315ce0a2e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/metrics.json": "d47ed86263b639270788702675e5d4b69cb72af0ed60605b27e73dda7285cb61",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/facts.json": "b4f0804944efabfe9b3de90bfe46a15a05f9cc5667795403b7f96be60aed18af",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T05-03-10-030Z-ktls/codex-inline-judge.json": "1d806dc1eb9fff69b86ae1f2bf1eb53b289db9204e708edda8dedae5f7d52dcf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/result.json": "3bcf43677c1750d1449fd8ee533915a33e439a29b4f7913095219c84a387a722",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/score.json": "a95118d4ff41128c4200b4f6f220f1417e891fcede78dd49074abe69f47e4389",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/metrics.json": "998b83c181fd658e18985c356279f9a24fae0ade3e29754921bb073881574961",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/facts.json": "dc92758167fec52acba4f556f41a8cf6946259bc350a1a8b5991eb3062d55224",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T05-41-17-655Z-hsme/codex-inline-judge.json": "bc0305eb42fdf4ff0d41d15f17790c22da6407c4c5aeb93f9d2dfe90a44cbf5d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/result.json": "a36c890ee957615b95464dfebfc877f812e40fd051f5c323937e8db8bd946ad2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/score.json": "b0b77aa8c88a59b363b2d50c3b320021fe574044cda52277697128bedfdf1a32",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/metrics.json": "42cd33dac3fe5fdece33b00fc834f2785efdb874a4fb406b520559cbcae5e104",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/facts.json": "7725d9937ddf6f009f9fde7e56432d8b65fe753659d8f53784e9d89257d5f504",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T05-49-15-844Z-y82n/codex-inline-judge.json": "a432ed150eff0421ea523f23c9e20ff76ec6fdd91844ce37e2acf46be6c32e87",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/result.json": "f610ebffaa1c9a5e7b3c395c14ee14df5665f013e93f8aee85fb12ff2a49e8cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/score.json": "947643bf4012cb634fa45a913b672d41f2d730917661a6314b7a5482d663aa9e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/metrics.json": "6fef8e6f3bb89504106d5f39277ec727d38d6c75fe5992e88c4b255ecdea2c45",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/facts.json": "3975d608add62dca72255d4b6387a58d7e20413ccc66f2b3aae8b5602eef3a7e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T05-53-43-473Z-qa8j/codex-inline-judge.json": "bf8181355255219d9bdbb99bcb4d369e2273e78a6248e4170826e8e638d6faaa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/result.json": "de094548bfcd77280d2faa022526c2b34635ba27a7981b59a72d8833d2c525cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/score.json": "f3d223dd61fe8c951f8e5647ca51e39a480dde66da118c647f8680eeeb61d40f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/metrics.json": "89d3bc3bec8ce7b6cfc266681621194e85b3fa7584577fd0f60df9c567a8abf7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/facts.json": "9462eaa7e2d0b3386f3d7d08c1006e320f22954138f7265d91fb74eca3362699",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T06-44-19-925Z-bybk/codex-inline-judge.json": "24e5fc0746d8f577402371a1655a6a8298d154018500c79613bedf3ed5e22dee",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/result.json": "8074060bd5a73a30ad8bc3286d9ce399f100f76e721928e925c69cce6a57c94c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/score.json": "6dfe61b4da78fc76e5bfd919de0a46ec7840d9806b3651c0714a46280946ddb2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/metrics.json": "c114f881872e97c224dc0a48b89eb4bd90a678d99f9cd75696d070a698a602f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/facts.json": "259a308c673534e6c3019d845314e15e3fda4c0a23078c89543cba9df39b5c24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T06-52-40-699Z-jnue/codex-inline-judge.json": "315e75989954cbd6387e8e4b77305faaa267a48ff8bd5383f48d4b163ac05f6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/result.json": "ac4895804e4fa8c71e07dbb4f0a808ff031c5f95a7061d04b46c17ea68ab0f06",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/score.json": "858c1fbcb59c1cca0eb1bb9aab68d7677c2bd375cb3604b0ab693055ff41fc6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/metrics.json": "99cde3ff20de44e277a36ae3be617f174985db95f471d124eafcf563801c2537",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/facts.json": "94271ac80079ff9f7890c9890446665f1b3fbb3883c0edb3997b562fab1d53b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T07-16-24-146Z-l13t/codex-inline-judge.json": "891bfe7c53915b80b89df94d95886f088f948cf1bf486522cd8cd94fa5756a5d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/result.json": "d7fc472b2ff08653cd6b57d9a87a1100dcc6d8a9fa35479bc1c20bc895a1b649",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/score.json": "ed0638b42615c704ed83c67f51fd4dbd8bfb6caee52a5579bcb2253e376f2d6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/metrics.json": "7dd896a6cdc8eeb2cf015d196d0a2a7c1641c10ea1f86d6f101098e29fc25349",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/facts.json": "922a839eb0c0879c7d63b73a6d043598924170c2021aa7c6ac8517b2cc7116f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T07-31-44-901Z-lxct/codex-inline-judge.json": "960c8fed30e5f4664776b1b8b5449b5daaa4c99d2710c0572d962f2ef75f730d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/result.json": "e45596b9a0c913ec224c4c71c4de196c6f0f46760e68d112ec66d5a44e8f3329",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/score.json": "a8801e148b1e837f874e05d8fe3a7e7f36c04a49a16fc4123a7313407370e051",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/metrics.json": "4727379d76cbf3d20792531fa4c392e11c48e0a1b32a1b83441f9060e36941b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/facts.json": "8bf15c778234b855667d6ca5ad761f7444ebd33e755328c635d5a3b1b47d0f2a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T09-18-34-283Z-xqxb/codex-inline-judge.json": "39fbbf6855119a727a01d407f4daa88f8d1a61c34e9c98119c2270f7eea4f1b1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/result.json": "c0be5847a1ea5265c34b7ca11dea2c48caf98d0d22d97e76b7599424d56c8dff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/score.json": "3aa6afe1a548e3f2a7553b97547a972faf7b89b1f022149a94afaf47497771d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/metrics.json": "41744fb0cc929721c558772433b5a543743ce0a8c455b16d9256d00a964c85cd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/facts.json": "6836dc06c59945d908e7d9017cfa3277c45862d14b9e1210d4df571cd0f98cb3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T09-50-11-979Z-zyge/codex-inline-judge.json": "660e19b405ae5f47879291bc330ce41f11491ba4aa6ba4bfae6cbc054cf50b26",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/result.json": "84096d447fe471424fd93c2d8f0d3fd02c05d808906dcb462572762c184c1b05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/score.json": "d1c9d764c35b1a36323647b3b2f6e80457e1726e80dc81a79f9e9b56637ded18",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/metrics.json": "2cd327bcf4d3243c439b7ae87071cc1276582a52c16097a471cf6a8362e3eb34",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/facts.json": "0c3696edb7085a44abff93b529d82256569ce75eac09cbd3f164acd2c03caeb3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r1-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T13-08-17-126Z-y95x/codex-inline-judge.json": "b1fe6f23fb22fb8d287bb7f6e7d6b1b2d704b6cc0679d478f71fcbab2e7c9d86",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/result.json": "325b3033b50538fa26104c49670adbbcf47450c42d2139a5bb945f39226effc9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/score.json": "442b4ac819d7b66faa9afc4347e45b017d53496cde7ddd1cd6f3acd162c4e359",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/metrics.json": "3a2ce8db55124698a5f2cff8f5bd58c87b06c009b8d456ec024296554929f5fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/facts.json": "40b68313d7a8a25f7c63cd720304250310011cd44cd248a6b2489ea33bedba33",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T14-00-43-799Z-izgt/codex-inline-judge.json": "83532f7cf6a06af6a92c83216c718acb61d794542fb3e17d09d4e529dbe64401",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/result.json": "09e79b37de999fae7905bb65b47f57fca47dd7d3ffb76398375ccfc38394084c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/score.json": "fd2076104b9cd3eed68c1b68fe7a77283ffea825ef1571ca61d7efe188069f05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/metrics.json": "74af18b6365b6d3bee6516f8220c887318c3b41a372c90acee00d1f2144adf45",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/facts.json": "76c4f01ac080840f7509a1101f55021abe74d2869a4e47d6cc0b6827c880f50a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T14-05-47-784Z-1rhh/codex-inline-judge.json": "909bafa75b92c0849335c32f8e969da9cb6b8c02e0d76450d8664c7d25c08a7e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/result.json": "af4ec3d5a13f2e73c5728f4517d1d96b3326111723a84ddb760c6ca3dcbc0d11",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/score.json": "1483e849f75f327b800445909b29b8cd145c56d970458f59cb2da31928d7be1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/metrics.json": "3528f315be24e008c08b4e9e6bde176852e097350fba8277d3310795d5c2c061",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/facts.json": "57202a5027f165a6784955c94d447e2c16e670bc006fbe694287cf889aa0cb6e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T14-11-58-422Z-5cwp/codex-inline-judge.json": "7083da5807122f59c282a3b3a9cb31ed9990405e2772d8b559789b3ff777b1a0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/result.json": "9f5694eb27fe0244b1a4c45abb49fac61bb27b5c53b53301160723abd7efab35",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/score.json": "8194b9866beda53dbfb5a2ea0d487399c9dcdee6381128c994f44ccebda35922",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/metrics.json": "38eff077decf1c1eb4558f8717d63a83131fda360bd67fd0eff26968c596c9ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/facts.json": "b6c5a2470a6fe2f42919b838f9f8ab851e1704114319cc98f16e39191243e550",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T14-17-58-477Z-j2mu/codex-inline-judge.json": "db8ceecd98f2a2a15c6058fdd632d81b49077965362db6d4a82c6df1366dd7d6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/result.json": "3b5f3bde3524920ef3f4d1b1cf7b368478b3289ec878e6fe39ffd0869f9bcefe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/score.json": "2daeaad7385a8a78856ee3231b10355a23fd119556b4671416445333be12c2a3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/metrics.json": "00a839982c772b545ab3f306ac6da0eb13fb518174e1fc4420eaaf3fb85dd272",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/facts.json": "707af2a3d9a6fc815e47e46fada2edf693bcefec232de0f5cb25300c61c1c36c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T14-23-13-871Z-estd/codex-inline-judge.json": "a3126eeafe2897f8749dabd6f0cfe762dfa52d6006e00ed1a80a1dbb03d3bdbe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/result.json": "b12b8571984d62f061e6351c87b2930df9448c207b2bbb7dc44ab7636bf7f050",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/score.json": "f6a90651d7f40bc57224c358929a23cc91cc6686b6491c3d0abeec0a3f368a7b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/metrics.json": "f94135197ced3eef91cfe5a05d0efc742e3af3838d09bd9ab48fbab03350fd6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/facts.json": "a6cf4bdb9f25c663d57b39024c05df26b6af02c0403996805de6286846e9d48c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T14-28-12-069Z-nooh/codex-inline-judge.json": "a6568ca9de877f88366c78edfaf28f21e2b54d54c0e5e977893a356421d0bc0d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/result.json": "5f86dbd87ffff3e80d566d4b0abbed62139d12a07afcf5b69bdc25f622b0ab0c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/score.json": "4d0eac710e65b681062e1bf605bd2207bd872912277e9ad6224732129837655f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/metrics.json": "f24bec7b5438078ae7f6cfd33b7be3ddb9b38960400d5e34b93e12d90fbc1a99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/facts.json": "4e72d355430e7e47447bc70b39bad549f0fc42d18d35fb2376671b798d87d98e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T14-33-07-793Z-t5r9/codex-inline-judge.json": "6c61384c839ea34df936c96493532d0e3b3458737dc5699ba3bfff065d638626",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/result.json": "3091375717c10a5c5ea76337f5d291177b4addc59bb45216130d96e939fdfece",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/score.json": "2634a651470b1e953b031bc3c77d859e4216885c1a6299255a3ea2f55785d9f2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/metrics.json": "7730221316c26266384c1463ce9b04e50c43fb8e3317e752ef0167776d576556",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/facts.json": "2763e26773d02ab22eb6c5915a9ccbf1f89c2b2a9dcb169fadc090022e23f6fb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-20T14-37-58-578Z-1ssg/codex-inline-judge.json": "a9cdaae6b0197183a72693a61863ec274f107281dd1641ce40551ea0a1d051a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/result.json": "e7ef4472d11970546b3d9b2697e941544e062e12145d13029b9ee378138a304b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/score.json": "1167ec67e5291856af87f0e73be21dd1e820e793374d11c839431f3c0d4bf144",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/metrics.json": "44c7bf48ebc915c05241b2d0927b5842296e74644d97e824c832627afd2f311e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/facts.json": "1b7d1370ccf0cb743d14c558c0eef736d73d3191efe599e4bdac53ac306e100b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-20T14-46-12-996Z-28lc/codex-inline-judge.json": "21dddd9205c29f51d9800f0e085bc4cd8ae9c9888f398ddfc56064f265d44572",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/result.json": "a50ac1a94204b4781977dd75d7970787f21779b29ed349a84be58d638d96ce8e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/score.json": "f9b0cfb485ef74568a6c6a1071f7980c854d3f1061eba2a511a81451c7c4f264",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/metrics.json": "153b98ed165df45d6462dbe26bf1feff941901fc04391b68f6a5d2b01b7b318a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/facts.json": "a44096b29e7f0c42344f8144d590a3ef9c8c0dd9927c25c196d65aea7c6618e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T14-52-03-408Z-oykm/codex-inline-judge.json": "5c04f59b4015e1b01889ee2ca473732304b4a0b3daad8f3b0ca091c7607297fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/result.json": "10902bc3234bfbc8c58bd47f8cf51dc7d05a20ba667faeb90415bf95eb303f63",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/score.json": "d6bb3d16d11d02d43cd98fcdf9e489aeb1abfba752e9b38b70b48fcfe8ffc943",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/metrics.json": "0239cea1ad911e12a12e4d5b3b79d1fc89bbe452f9217c0f6e05afd9f55eb1e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/facts.json": "cd4e5ade37c61ab36bf3cb7e08fe168631dc5848a80e8520ff91ab940271a162",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T14-56-55-576Z-hp4k/codex-inline-judge.json": "8039c0b240f35f66424d1ee16d382ae1f44d0b198039e3d8b323db76f058e8d5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/result.json": "8d66f7e9d1a070a3aa34abfa248a352c93a1d9038c6b9413ec59f829e7dde468",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/score.json": "75eb4f13b9f16ed32610d478515c15ba6fc8b42f77fc2098c05d3b3f86bf88ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/metrics.json": "ff2965d61ddea1595ca22f949f0e1bcb9ff4265eeb7162953e05d77cd5ea9a77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/facts.json": "f13b800147714a3858d969d9fc33684e31ed2ca15a46388fddbcb6d1bc7d2404",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-20T15-03-01-617Z-ber0/codex-inline-judge.json": "1da0f7c40cabc1f6081ffdce368c8b5968774e039027e6c2a001131904429b3b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/result.json": "6344fad211a61c52cdc226d6096968c149de126428091531214128581f35c528",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/score.json": "dbdb3effeaf082eace151a6f6c794a8380679ed5d650163f44091cf9ef87ee70",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/metrics.json": "7e9e1fd1b8af8bbaf32f7b7a92b222cda43fcb3140a6fa765cc8cd052ca05ff6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/facts.json": "82f39c9a934d72729dd5dd0de5f0d89a2475e1e8e203d5dbae030020507fb2eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-20T15-08-01-333Z-83v5/codex-inline-judge.json": "f5f6e7b45dd97f5737ab1aed09abd636d3851ca2c405314d20d01ebe504a5e9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/result.json": "5367f93bc10f1b7d5d6307652e020e27dcfe685da43b67d82c8f3814b7ea0b6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/score.json": "e807bfdf9f039e7b2b045c061417bf1f5a88f3449f41ab7f3b8cf5806ea40da7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/metrics.json": "2e0190d5770abf98cc65aa52972c1eb4da1cf52ecaecf9200d379dcb30d4e08d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/facts.json": "59d861f64ce20a93b683ff26e2e437589700fa14790759082a2f8a6d35792699",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-20T15-14-30-934Z-sckl/codex-inline-judge.json": "f52f104824e8ed325bfdf468745139cbc1f81d52fa0aa1aa90bcaf09e45a6149",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/result.json": "f0a7de363473800e02014f229ff9f3479213988b2fc8267628c41310b4d6657f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/score.json": "7c5b8e96ecd1b2490e56f54a0f63ac9e33c0ef6b34a1d52c83bd9dabc61cd53a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/metrics.json": "180f1e1e9d387a40e4efc96d9dfcbc6f0e6a9e5f7679f2ba688196f9f1713f0d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/facts.json": "896a2976123cb768fa3ca4baea2946285004632edf2f980828185c2f30c43f9f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-20T15-19-04-405Z-s54t/codex-inline-judge.json": "8b1fb9c5aeb6f175640fd3436cd01a9b9a6e9d19e3b340976b605c81f773bdcb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/result.json": "98ca48f6e228cb858bc8ab9abd757dab745b2e50dd5ec9405b04f165cd4e1be1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/score.json": "f58644de6618861c3ad347c22e53e740fafeb78c77128d341ce9ddde4455babe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/metrics.json": "0d16da3bde033d3c27bf5348cfef29ad6c3403fba083465ea849859eb6535838",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/facts.json": "0ff06d3c873bcde93838a9dd77ac09567c5aa2b7f340246bf128ad2072bee1ac",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-20T15-25-02-898Z-gv9w/codex-inline-judge.json": "0e3b7335786f07fddab181d9c9941a4a5718bdef248a9f58442412217cf0a189",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/result.json": "5044cfa3b718993b8c2087c97825e436c6623fbf5805582f7ccd7ffc3b802ff4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/score.json": "6bb7880f35eb93cbf83d4588972c646ed2e85523268d3303e15fae52577ed7fb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/metrics.json": "c433ace6beb62d617cfd90c2779be014f3a8afc06295b5a512ecb39219cc64fa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/facts.json": "43adda62cc95608a5b097dd7bd1e1905f0d220913db9a84ab8f8e0b25a95e312",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-20T15-32-00-461Z-gnhd/codex-inline-judge.json": "11f4662d5ef866db1a9206cae985bd086f54e2574c413042a2d8e5fae17c47aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/result.json": "7b44ccb100e8b9582e1ac572543f59a90272641ccfb9f28fa38f6b819d555be6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/score.json": "3ca8c025730b859217efbb7796b46cea9e3a3fbcaaf7badf2cf7d6a33c877b5d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/metrics.json": "14f0e624152292d51f6da91b06f77669d98be1145056823eec8180dc494028d5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/facts.json": "79ce6ad36ef0b434244049501ca76deff1160d41b20e46697b2c2d0fce5ff562",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-20T15-38-58-540Z-q2pa/codex-inline-judge.json": "429fefe24cfde3f9ff549d1b44e861e4b2af0c8aded753552abc3334e94a4d92",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/result.json": "6dc57de1e6626186b2d4856e0996bac5a6b7524a8fe5c6c2ee0c722863eb7b7e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/score.json": "d5c734f6d2f01f9aa29e829252d678b4beb0205528d67f9416f9442ab5d5aa89",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/metrics.json": "bb1b13de7c65ba9f16692af82d2c6ef62e73c1cfee883ff32d6af4d54feeb974",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/facts.json": "059ccc149d5e6fb31fa164a1c9281521c75667f11950739e3d332c2d3d04e14b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-20T15-43-29-024Z-1nvz/codex-inline-judge.json": "07b4f7451d4f905f5b9a85f761e15a6e09fcbc3c150601f99e450aa67de68e00",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/result.json": "2fb61cb1e0dcfbcc099080872429d5d37c1925b8564671edecf739121bf9db90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/score.json": "d455f75f2fe46b4c0d42a36dcf749032ee5c83afbede7b5814e01ce2d9f17ffc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/metrics.json": "63df45f7b20a6d94539a7ed9e8e9f89df9a0d4ab79dc64834ff5c6165311e70f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/facts.json": "e0d585225953b535a82afcb47f24274a9c68b7f75a81942419424efa750e0914",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-20T15-49-30-364Z-cdbe/codex-inline-judge.json": "23efa7ec2f6b24aa0092245806b327cc73cf6e0bf19642bc60569adf3ad03773",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/result.json": "4305e3205d3ad7ed8c986b7d6e9b4dfc9cb723fc570092390699aedcf8985c2b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/score.json": "8096dad5bf9b41943a09846634a49bd066552465dea26fab94f9166504491c1f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/metrics.json": "6ee5f671b8551ba1b2bee18a7ff5e6698c4a0f103b166a0063a1535e58c9e3e4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/facts.json": "dffd9f8a41011a85058016d5c285a92038b7fa7593f9f80f97a143593068b759",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-20T15-58-10-699Z-xxlv/codex-inline-judge.json": "c0c23a5324856694be6527875ee2c32b00f56b99a446f61b78738379ada8b578",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/result.json": "bb5ee338ab8d72dcc657a305d8a2b7ac3a278b4d9460a3f464ae3289669ac5f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/score.json": "08ec85d4d19924693594471b68c83548d543fe4ebce2bb025c2968473dc1a311",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/metrics.json": "1dec8cd2389967947ad4cd66284a21caff4384aa53dd9145e6d373c6e8c04cb2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/facts.json": "74ecf9b8bf80c966603f9839ca3de2816093b7e308ab6b35e395a21baf479ea2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-20T17-32-28-869Z-puzm/codex-inline-judge.json": "f3bbf32ccc084e64f7ed773078fc1db76a129c5f1085450b1ee0a73c932798c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/result.json": "c291de117cdd989b8bfa32a63caf52ca86fc2fad0197386bde50589fa8c53cc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/score.json": "2955a7589750249ad496c691ba9f9ec43ee8ac1a71693f5c9a3c993777971feb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/metrics.json": "57dc965fee48ab7a7d822e2b1ee7d206a42784600187c5f4e3361ff630a2f5af",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/facts.json": "c177b4476196cd852c01bce1d4ea0b40451f11d3ec4df3936efc6045a17a767e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-20T17-41-49-894Z-ly5m/codex-inline-judge.json": "f766365a72c52655e1543fba9163da2bc0b6c1a402777597e1486fb426f9cd66",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/result.json": "8d4fecf7abeae355cc51d70657fc5e89d0c3fcba40d8ea3c51654b6bca0dbebd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/score.json": "95bd294d5f560ddefbe6bd7d7b3c146a210d314541e6d46d366d2c987f297270",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/metrics.json": "87400575fb17398b5750e8bb89ad1eb0fbb757d904697410a2abafb5b7924187",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/facts.json": "00e306f2a25f46925b90941965ba88841a11fb8d9b4f241876f9602733737c6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-20T17-48-04-549Z-smdm/codex-inline-judge.json": "10d869befb0c520c87c7562fc1de5c0f313ee15572dfe90d4c5cce1bd7d4730b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/result.json": "ad057c632f577c156c6f6296c943c1e9ad521cc66c591278c261b8aad6843d15",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/score.json": "62e2a3d3b094a5002f24b2a364673d9979f7824207a52cdff56f92e81b3c622a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/metrics.json": "bfd2b98f6121893331ce22b22737200f73faad4b2ee4225812dc6af01d94eedd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/facts.json": "c33f7fb005f9e6567df8d51fecd5282d8ca9fc559cbde5de79fddb8e5d3d34d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-20T17-54-21-281Z-xt5z/codex-inline-judge.json": "498c4b5fc7d9cf6607ac6f7cba60f02530f3823b0060a08101022815ebdeb7a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/result.json": "e6a3cbbdd84ca7357b948d334244f6a82d0e1d17b8d1aeca2def78cc67f40c10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/score.json": "dd248d5c8639025730b5daebd86e082b011f7acfc8260ef33cd6ff9a6d66341d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/metrics.json": "69b0bb696b36543d13384a5d9ad5484e17681f227576b9b8e32a32d36fb67826",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/facts.json": "4b84205ee279cab0ff4335a563e6aac39c654069fce6bb3ac914b57504de1bdd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-20T17-59-28-093Z-zyud/codex-inline-judge.json": "3ba295876c9dbe98bc8d5423149389e842794a261b8c5ae184eea6801cddf375",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/result.json": "86519e47aff8095145f35523bf09e8a2289f3909f15a217100214a2b29ef7f8f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/score.json": "09ea703fbbc8cf072822f9b0e96faa8e8a7ec9a6d9c90e258bf711a1fc14afc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/metrics.json": "d0ca0de8df8a4687931ec12fa0b7c2bd56625a21d43d83f6a623b081753e7851",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/facts.json": "c8e44433836840d03fd824d6e8e471f18b2ade5781764c8af23d96417184f388",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-20T18-03-54-146Z-67ae/codex-inline-judge.json": "ec7c73bd8ae266e4eb585946c91ddf1a761de617af153449161f7a111d83614e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/result.json": "f688b7c1087c0a467be9bfc8f0a940f62d69cca1b7d5c8b85bde19c34c7907ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/score.json": "ed384f83a13e5a1f6583f75cc1b71e6f8918836ffb51cbaca5c93934a39129b5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/metrics.json": "c7c171200466134a7b8ed1ad0394b80ed0e83e39cab4d9206b7f96f019c2e195",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/facts.json": "f063c6dcac88ab748b04b3dbd836deaac0716637c23a6db19288fa8a4618e7ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-20T18-08-29-491Z-rtdb/codex-inline-judge.json": "8159f838972182308b30ffc94d4d76ee495a380c153f5acf88f0e07eee43f505",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/result.json": "7def1953ed6812d8b4a573585a07253d41f60fa354ce346d3fc9a67d86b80fb6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/score.json": "e9065e01cb703b5301f805c5794ba47209579d73b89cc7cd998f240720b05a46",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/metrics.json": "a48db63f5a3d2cef27eab702ff964021528720408e906e7219f7e7822cbfc62f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/facts.json": "7dbb17dccbef7cda23fceadb66316fc650c7e19af65fbdf1e2cee71033d86030",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-20T18-13-45-577Z-hvsl/codex-inline-judge.json": "8b6fb70b9d43b941ffed5a60f893158bfe02474a6c8fde56997b660a7fadae17",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/result.json": "d9093a1d9a01f7700f88439204f404415a9ea0f1fefd4cd774f9303ed12fca70",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/score.json": "70e35c4f940b52bf3b89b5a1b88c5a4f3e04bf62bf5ebc62ca3f7be163e0dd46",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/metrics.json": "79f4f44303e103c9b2da15960e9197c3e69bb2f1ef5a774088cfa6866aa5e2b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/facts.json": "08367d9975e37dc9f74e94f1284c84da47cc2a708202661e5e563f1eefb8f283",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-20T18-26-04-330Z-5x55/codex-inline-judge.json": "f2db48949bb84261331e89fc5c84c449bf7bd031e708ad9e71d532b3d5e1c499",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/result.json": "8caea93a6f37d948b44ad10608f4e4bd0c633ffc38a0dd98e3f2a00f3e0d7d14",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/score.json": "c0b6ae3abfec567ba835a43f6994588f821bef3c2bbbf5157212eff816857d17",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/metrics.json": "3fe7c14f5b5d47ee2f9cc4b36aa9a188724f23a6d846b2d667d349abddb28611",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/facts.json": "f1cd9b62d47a44118d4d7df9220fab6f0935c2e8d7559b55ea4d266fe3177481",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-20T18-31-33-210Z-fl45/codex-inline-judge.json": "29f52244d5b856a8037a3a976d6981bacd40e5c41baff19bcff429be212a4f23",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/result.json": "382e651059c63a1b475c28a3ca757e36163e7f0625e8613d325bd8b8b7f8530c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/score.json": "e0076ed98433fcf50afd662c62b087e451c29ac2689a39fbe60f54baa73b9387",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/metrics.json": "ef772732d0bac1097d5ca720eee5f241a115003111f785a235c72e4703feb0fe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/facts.json": "66016cda4b40aa5df3bf19908c28f910faeba09a48fb722a6ab79ddbc2b6a5ce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-20T18-52-02-199Z-f195/codex-inline-judge.json": "241dfc02bc3dc963218a726e1fd636446ffa54f42f5346bb513ceb3ff94b3de4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/result.json": "1fe126ecb78100dcb77656c2165c067067f0cb13cd23fc6dffe91ae07d7f17d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/score.json": "98dfe7f4449c072ca86b2edbc4fb7f75bf34c6a23b4912a79300ac806c875e02",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/metrics.json": "c2a332eaa06a41a70f3bace5ee95614a225d7e3d24facb0c2e8f83a7e65f907a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/facts.json": "1bab9d6b3402d44972d83436d73025c579752fa2a50702a81102581bf0cac9e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-20T18-59-40-380Z-ejji/codex-inline-judge.json": "851868430f68fd4733ba087141342038c5d019366c1cac829744347b713f328d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/result.json": "815a168738c1ca7c649f2c1b0db224067f981542e57f201cac34b2e9df3bb59e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/score.json": "073a09ed95043f37cd24faff29d9cbe2b4b11411ccaa217d60c6ad26868063ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/metrics.json": "1338c2b88bee227539e86b4ed840ce4d192c2acebd11ed319c43e50166e9c630",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/facts.json": "f676777eec62842da3c82ec0efcd68c1b247c2f24477dffb0858870d3d2aac3e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-20T19-29-39-300Z-266m/codex-inline-judge.json": "2fb96a59440e30f9543847dfb84690fe6ff0ff7f0d4ed59c0ac37018c1d44833",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/result.json": "e675ed75e7c71900585b5a8e44d38e922716dc92baac0d23266fa9480258d178",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/score.json": "aec98ba00df88699e32eb531ca48e98ea1a8f321d51af7e68c908d652e2d715b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/metrics.json": "561acfa19553afb2c6f511effe5bcee2317d3505dad1a08212ab9fbc20578fd8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/facts.json": "5fa4ddad9ba444c0cd94ccbe7190a8f97b8df1aba7968975c137ef69592f49a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-20T19-42-05-979Z-cc0u/codex-inline-judge.json": "dfa9fcf48b67348917e710a18c979df259f90f7e5584c410a91c9349366eaac2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/result.json": "ba99ab2ae9d86f79010789d43da4265e15c238edc592a6dcc80afd976f3f5a37",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/score.json": "a134d02497979d0e2fc0453ef918eaa053fe1503391eb80e8bbcf337aa391ab0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/metrics.json": "435640265f5d8abb12db07935dacf7e611a6014a6d21c6d1898933467c116335",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/facts.json": "c840aec89d3af94dc07ba94f17697458bc6abb7872a279345ae68c91bf5e81a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-20T19-46-35-714Z-m24i/codex-inline-judge.json": "0b2c3a384f4a39763ccf164e43c613d833cdd23de2155b1d869ddd297eba34c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/result.json": "58d26cae178d98aabd36915d144ef08f9deb9bffde729a40125ca735b36c943a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/score.json": "881dcb53a05f3e5d21f8fd2901f71acc25ea992c8580bfbbd90440960d974861",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/metrics.json": "1ea00872f0e0c64071f2c8885e6a7846806bb8506415af897e9827a06e8ebbf6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/facts.json": "3172d4232d7c6b04f5aa2060cdfbf6e9dfbffca222a01c95c164db8b863adffa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-20T20-42-53-951Z-r44x/codex-inline-judge.json": "561985f31f554537a85a55588958cecec890fd7213c54768db05b28d3a9de91e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/result.json": "5fc5eee80f248daaf1c074821944c84dd8f8d1b7d2949dc1d28ea3435bc76376",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/score.json": "d3edf7254ead4e6a336d782c8ce047331b9df6e59b4037aac2825dd29aabf042",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/metrics.json": "c97565aeeefd6d93b321b19ceada3140dc2a469c20e878ff4f819d7a8383aee0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/facts.json": "d3b4c96da10a7223b5b0da17f8fd65ea7e9a7e485835f8968c7d991ba97b8799",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-20T20-51-39-350Z-bf36/codex-inline-judge.json": "9ec7353fa5750f4c97dfdcdcccda101ef220130c3fea4d0f27cca93c2e4913f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/result.json": "06ddbb0e0fd77dabadf0731bef2e2aa6d8ba95450f036861c29b591492a217c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/score.json": "1c374caffbeed191e83eb83d38f889a442a3e73998b22bfaac39f44d611c8ee5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/metrics.json": "e1ffed06ccf5e97a5db14cb9ffe2b4dee661f90760b60805deed7d325af0374c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/facts.json": "3edc1ab2d3921c6c306cc1a12ceb31c9a87c19b3e45a4393f7202ff8e599f784",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-20T21-18-04-229Z-ralz/codex-inline-judge.json": "62a7808e9f0d83a004d0b689125980301db9fb1e27764a676afbd8991b7deee5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/result.json": "b0ecfd244aa1d598177275224174c2d2119d8ca593e8e69b904a89d76fc32b6e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/score.json": "9f6a94d6d75a9de7603799d0d6bfd13057525d77827abc17dcdde34764006949",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/metrics.json": "a6cafbe76b89810e6bbd3d96ae80051475066f0a47a7de9ced5af6d9cadd337a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/facts.json": "cba3b7323257d9516f8690c4a159d39d5569389805da76168c67da201552f974",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-20T21-35-24-529Z-f97w/codex-inline-judge.json": "7addce6cba61b058f8329abf6589f3e4635640506efa3d579aca42a12f28ad38",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/result.json": "468c705c3ca3ef900f161b3cb9be60f17d89539842cd3174f2dc58ce2ff6aebe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/score.json": "3f0bb062b0ea9c961537e0f494ea9064864980e47489f270213747d6c0717ea2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/metrics.json": "7289fd4446fc474dd33eeb4b7751c79919c39f343d0a35a157a354acf239f2fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/facts.json": "4402f6e90e53cec4331868bd4db49e6b0d2dda1525ffab62f3c6ca81f4108189",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-20T22-02-15-867Z-vodk/codex-inline-judge.json": "5c79d88144f37bb6757d5d1342c38e13f0c954249477a85466ff76c218787273",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/result.json": "716e1816e043fb979d7d3741f91fd6f080cfde8e24066c34c897f424556b38aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/score.json": "c860a6fa3a798e740b0b50530940f18c20141d1c0adb11f4aecffb38fda68810",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/metrics.json": "2eda13ea39b13706398cd4892498220cebd07a689dc12fcd75b31b33a388f57c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/facts.json": "8059c005c0c4696cc8be0b183f03514110198ca88238dbeb73431998ecfdb29a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-20T22-33-28-655Z-jqjw/codex-inline-judge.json": "2386b1fd2e9ad6a6ecc9b42947b031e3253909a66bc3d2bdefda0c973111e433",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/result.json": "79242a124c9ffb180fe1dcd8f4e40649a07673704667616c0ace31b0dbf7c453",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/score.json": "59da607ddd8b83247530137054518026eda457d6713e9ad6b476e1b2421ea7e6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/metrics.json": "122b2c9aff789d66da9a04cd790d2ef53ce0ddee6576d6dbb0bc84e176c7dc22",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/facts.json": "1ffea674917a0082f92f19d490bfe200bdcb09c04ae4f4e43cae4af39c7076ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r2-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-20T22-56-32-308Z-r17r/codex-inline-judge.json": "eed800f6c70ff14d13bc392907816a76bdd75c30818aa50b1230ec5a613bc3a4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/result.json": "55e7fdfda88052f68452e183e9eff1a897dd38b1eb2178be5d204134495be992",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/score.json": "e68c355b001fa23ecba2b1230deb864fa0b1173ca7c829213a658345afdc175b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/metrics.json": "37877501c9ba980059893d32e4ce7d4dc3a8e7ef3dd9ec9dd560ec02371e02a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/facts.json": "034654fa9cd6679478d898c2cd49fe5de18588e076004e4badbf69d0faaafd18",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tictactoe/tictactoe/tictactoe-btl4-35b-q4-2026-10-20T23-23-15-684Z-k3bk/codex-inline-judge.json": "09cba432e0ecf118af609546e075c80a618704d67eea6eefa69b471ff6d25a8b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/result.json": "39ff286297c5104dd403dbf488249370afb28c1a734562529c78a50c954b27a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/score.json": "51b8944952bdb0ad24013e911d0baa4b7c2e01073b9aeaf801614d658efdc026",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/metrics.json": "6447d2c8ef26d47c02b68888c47abfb7f0f936cd7444178d2c925671161acb61",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/facts.json": "d25d161aead3008488487246b70f5a3e52ad905373cfd821eae0b38e708faf13",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-petshop/petshop/petshop-btl4-35b-q4-2026-10-20T23-29-14-621Z-d3bu/codex-inline-judge.json": "9d9b7a77f074a6bcae1c413fa128c06f8ed8d8db836aed04e99f81d6d82ff61a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/result.json": "8bdab3b1447cef54355071a824c1609896b5383511d0aba92f2e87cd91cefddc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/score.json": "6b56a5b2ffc6b0a118bcd89b32d0aaabc719b3b7d467dcb0055e0f5c91ed5ba9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/metrics.json": "6a25f0501ea1d288af42def2be2223dd1ef13d24f2bcc322c3819df8b126b2c5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/facts.json": "77a5f6a39fb0ce23ffe0fcd8a4b4620006436d5295f9cc0ec480005825c680c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-tankcombat/tankcombat/tankcombat-btl4-35b-q4-2026-10-20T23-36-00-615Z-cdnb/codex-inline-judge.json": "a120e369668198f0529077d9baa1cb5e7218a600f0467116c567a7e09b9fa6e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/result.json": "371fab0f1db17c0283ee3a8b6bc1e2b057b770434b5635f12329057c40a87f4e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/score.json": "789168c8c7cb1f64ebc07a01bf2f6f96299280213e04824f17b7d174b20fb92b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/metrics.json": "df493ba4df36d56e00fc5f640084f7edc3e03001605a99925080b1cf6024ba51",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/facts.json": "a4694178eae4754edbc823e33f0b89c2a79a31cf366c698150138a039430a1d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-schema-migration/schema-migration/schema-migration-btl4-35b-q4-2026-10-20T23-42-32-066Z-1u4e/codex-inline-judge.json": "ba04b54aa1c6e910df83a664a3d0d3e5fb271d3e44923f7e4391d10525a779ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/result.json": "f6439ba85459530958c91d981d3a977b2b93873d6d7ee5bea0100f9f1a3c5b4f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/score.json": "bf4679fd50ce96ea159f02cc68f22cab73718da7112db3374d66c8bc80c30998",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/metrics.json": "bf11fb82890539cbc20bc379bb730810d12812645b6950511ae1e0ec3d14c1f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/facts.json": "7128d53a8c585d5ce1045597c8048fc545b79994471ff7ae2cf4ce1361e4cbed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-btl4-35b-q4-2026-10-20T23-48-01-372Z-a28j/codex-inline-judge.json": "2603d4ad0abac1112852b6e9898633d49fb67f38bb633c0ce291e6facdab583b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/result.json": "8afba7863aca6e3c76f1a3c0117c47d2f764b2152b82b8db47ad9abb47f394b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/score.json": "209f375685b9ed4de92594c0c21c88b4f29b70832cf4ef4f0618d88159d33896",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/metrics.json": "a4a65167640cbd65d73c5856fc2ee40c047a21e5e802b65b9e05063bbe9b62c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/facts.json": "d5a1b094bc70cf71436c71c04a7978250feba04f1780f9b27d5212adf589af2a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-symptom-debug/symptom-debug/symptom-debug-btl4-35b-q4-2026-10-20T23-52-55-109Z-zmif/codex-inline-judge.json": "cccc27386935d49c53960e69621bd2d7d2fb4c2de0e352914675b3f8965f3c84",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/result.json": "e16c3f6b1279eec1782a5dda081cd0b17260a3e438a007ed7b64e074306b386b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/score.json": "8d1b12d7b86319cbdbd4a1f01fedd468a8d44198e844702c23659bdd5d229501",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/metrics.json": "89508f0f73bd5692429e0d0f36152a59f9320f0361a7cdae665527991474cde6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/facts.json": "70ea9346e97903afe8dc54414b361e785b3022f48889c90a0ba478e85148c7eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-data-wrangle/data-wrangle/data-wrangle-btl4-35b-q4-2026-10-20T23-57-21-432Z-kf5l/codex-inline-judge.json": "2c332e439a7619e0289fe2c2d34934f8d4c782e589c449eef5188aba742a9dcc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/result.json": "9f02a966942178b23a7d0b9b25b6b3fed0e45cbd81b0893677e321e90150b19e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/score.json": "5724de02ecfc973d387c87b0291b2f07979087733a281ad15a6cdeac80b30709",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/metrics.json": "3b8b75f1b606f251aa386e5d144a6bb37299e7191787b1b267bb9eb954f04493",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/facts.json": "afc81e2d387f28170b6613e401377b68f894aa0abbff0d52aef8d082b44fbeaf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-btl4-35b-q4-2026-10-21T00-02-14-447Z-4in4/codex-inline-judge.json": "778177151259fb0426e2fa8820ce33212b0818369647ae4ea3d11d5b0135af4f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/result.json": "5bf7d7716ee54d44442ef7e298e4d64d8bfea05759d0c7580e271f2065f11630",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/score.json": "f46b763fca9cf35a4a2941f65c62352f0efae96c10a7a81ee9dcf2ce34b70a2c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/metrics.json": "21fda539ffcc8f48bf3998698375602a057006e6e18510246eed0818ce3ca78d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/facts.json": "dc5186cf59bb95f223377972a9a9629d032496bf640f62abe5d44fa48bd8f08e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-btl4-35b-q4-2026-10-21T00-09-47-764Z-qprf/codex-inline-judge.json": "3ee1c1eb6850c076de8f5284f13af474f73b3883036d46af83ef41fd84fb7f77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/result.json": "077a7b98422ebb7206b8b14bad6477e821b69f8d4109f198f7a9371444167642",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/score.json": "d591b9279aa292f919ac3da26bd0c4a130d460388153031f5d477684dde1bb34",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/metrics.json": "9940ab3286646d207393d144df8a9af72f58e119b003a6a0d3755f118405621c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/facts.json": "c2c19b67096f0205ff236730b7194c78a5889243333f5e6ece75a80ef4dbbe6b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-14-45-932Z-oqsx/codex-inline-judge.json": "f3403dd3cd84b21519df874569c8d73c445ec5875148642e09587092138651b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/result.json": "f9fda1d3daeaa43ed37526270c9e89cdf42d35f34850dda65f65d52b0688773d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/score.json": "b8ea97c453822ab2bd35b00ba09aa2ef6f42d0e141f887ab60aaaa2f02b8c937",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/metrics.json": "855e1df34af6f30c31d8be1705287377fe87877f56cca967f0814cc16cd266f2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/facts.json": "ae34a40886451fdaee3f840dce0a2502514cb1cc434c81df166894bde0790ef1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T00-19-15-681Z-4ui7/codex-inline-judge.json": "078a34e355479d1eb734a941bfef11195b1f9f1c45035a471a80d792938a395b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/result.json": "5473805756f7cb7727ea1ad405ca54aa1ebff3ab4775278eeb7a7519888f5dbf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/score.json": "34157616129b55597d4bc55d6676389b545979502cfe8493d80c4a8a9311b03f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/metrics.json": "6d6ab8438221f97f655ad0c696ca575e9bcd9288698d54a6174414de569737aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/facts.json": "6f6fd659a5d28d92a2ea25b513c87385e69952db1eec39827690345b1a18b803",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-btl4-35b-q4-2026-10-21T00-26-05-458Z-se10/codex-inline-judge.json": "20598335d3f087ee719b1c4a390448a6772b826d057a71b6767eb8a94f6f23ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/result.json": "8de5ab6b0142db5eabf6970ef8652254a09f373d92d2a7add7d80e255af1bc38",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/score.json": "c560d3c4fb26fc306c6a44269dde9ddebe3bffa5aed1991462ac76827f01b4de",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/metrics.json": "803f49d741efd7e4565fe82ff46ca71fc9c72bac90389d073672679326a084cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/facts.json": "6c5c3654faa3ea8497d7de7c24a8a02e4e5a8ec9bf5985ade9f7586d971951fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-records-intake/records-intake/records-intake-btl4-35b-q4-2026-10-21T00-30-43-307Z-hg55/codex-inline-judge.json": "c235663b823f6032964e9862b3834231507cf12cbe3f263159988055c6541839",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/result.json": "d3f2657741c6c32c0676b9ccbb923acceffbed411c0f9f7a645ff0a1e3e7b868",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/score.json": "9223c898492b6555241cadd9564ea8323e464cc3b83ef2235256f11a1409bd47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/metrics.json": "ce7ffd91de25679ad2078aca50da414d974366fb390d73062d68139376e0298e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/facts.json": "e00b4602f283ffd902742149d5817572fd29e9ccdf1841d70ffd975beb05954a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-btl4-35b-q4-2026-10-21T00-38-10-674Z-e8da/codex-inline-judge.json": "3a51c9058b9ecb4bfab08d1c165ba218790cc677efc2eff22ea000a96d19ad6d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/result.json": "472576d861b336ccef5c5c74bee5da2309b6afa7832d78e95e0121ee1b8de4e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/score.json": "d6cc00bb18c778f4ca31d8fb9ea698f187d12233120b52586dc8d0f4b12b28f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/metrics.json": "d90dddef80f5e53a9cdaeee12dd001c96fa3dc8e59c9777ea1e1f1d203214c8f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/facts.json": "22ad63917433d21ac87e5a0559b766b4d21c0ec3d7778083a9abd8e892f9e94d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-btl4-35b-q4-2026-10-21T00-42-47-710Z-gphw/codex-inline-judge.json": "921e46bb9f0950ff7877dccdfc353eb83f0e0959a629e910d848af4c060c82a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/result.json": "b022191746b771ef129d07fb2d459f439ab6adcbce1d49ffa5e8cbee968707e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/score.json": "98b72ab42bff3a270f0e1519c90e5b135e217c9806fd0b3692e7ceb174ed6f77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/metrics.json": "4612273c13986308399c20976789efc492b0b7533df41369fab7db75789c3d43",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/facts.json": "45221d75716520e2914aae871c6eb16500b8368d53d8da63a81169d6c7265003",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-btl4-35b-q4-2026-10-21T00-49-21-709Z-ci89/codex-inline-judge.json": "1aa72a40c7f90d33137bca393fb03d4cc6651b1531da462a9ebf2cca2b009959",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/result.json": "311137f1a7a8b1a6148ce8464202ccdaf08f10e92c7b175344e6a40db78869b2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/score.json": "a3f6e311044873dc8580b133fbb2f3bcb661704cbb8473c1c54af46ffcba8d82",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/metrics.json": "a5a87f157909296cd05b00bb2d6e3ced31e2604e54b4b9f95d405eba5c3c5a52",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/facts.json": "b5035be8a2ac3b9a81db7997d5102f4fdab5a729f7f965376ff8ef5a6f82d17c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-btl4-35b-q4-2026-10-21T01-01-20-536Z-umm1/codex-inline-judge.json": "f73bd37eb5137378e7c723058f1386b38862872793abe5bd06bfdf6aa73c8372",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/result.json": "964f4f707782376dcbde49e4cc8d6b9cdde22ce769576a223bffb780865950fb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/score.json": "94a40a78098e8616eb5ed4205cd85362127af786830d3393ae59bb0233d2acea",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/metrics.json": "bc19c90b3146e03056f72c141aa3e5d860f92c5d969cb5e87836efbb5b4cae6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/facts.json": "9a36005e4cbe48f5bff02fd6ffaffaa43b50b15c437a0a2b36238e604080a0eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-btl4-35b-q4-2026-10-21T01-07-18-395Z-uyon/codex-inline-judge.json": "236520eec74b2178e9df5c94bdb0713838461ac60648267825670a558bf8ef05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/result.json": "299db2dae7c5f29ed5e0080609fc251c7b3c5e2e845b79110612ae47b2179233",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/score.json": "6de5c111af277970d99473387e88313349068990d8172904db60495ff4b464d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/metrics.json": "9a85f85da301bd1d78a7ba4da3e83ac5234f0b1e2c8aa905e6ada74692e68c09",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/facts.json": "b9563d7109a10568f689b5c573931460b62661547882db19cb9a0174d6804aa1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-btl4-35b-q4-2026-10-21T01-11-47-852Z-f939/codex-inline-judge.json": "ef1a3157101bdbf4d93f1b7b42e1d5225e154f55fca83d1113b6ab9e8e976932",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/result.json": "c85de5efe4452ddfb96a10fcb4605955a72f3d36c7126280582c87f7640f3712",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/score.json": "5a6f8e248e6666b3549352e2c98ccae917415c7db00eb868e92f6efb199ad23f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/metrics.json": "2031ab3a7dafdbd6cdd599f4f2981bfcce40d30913b0d07a20d74a68b465ebb4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/facts.json": "dcaa10910233b04434e090d7ed4a02d75161cb7d83987dfd72c335d7ca34daad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-btl4-35b-q4-2026-10-21T01-18-54-867Z-qgyt/codex-inline-judge.json": "d9b824654e21df660606e3216e85147aeff4cf372ccc962a5ed8db55aba186f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/result.json": "588c7cba56b41a0a2dfa9b424fcb54291658cd4ca34b7d70e4abff6e712f577c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/score.json": "0e593122f793f96fd544916c03a62878e3f1293ee7269608bbd31e97bea7f060",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/metrics.json": "f12c254f07136814d5943b0c359e014cad489be2cf8a65132f4c215121fd02df",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/facts.json": "ab277d71af1d7737d029e15ed3d5ecf7bba32239f1b8c2d7e3a15acea500e21d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-btl4-35b-q4-2026-10-21T01-31-39-106Z-q9n4/codex-inline-judge.json": "119dec40b021e189b1eb6d1b48474f5df92933a77d3a92a9f77347d4726ac8f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/result.json": "d501789a5f6e0bd09c021252c54c6eabf71fc2bf0fe03b8b240819f5e4d60928",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/score.json": "257d99a12dd7e12036b9b997433297bd49ecf1a9d160d538c27b2cf052eb09a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/metrics.json": "c9cadc5dad8312ce709f2dc903ceafbfd130625d3d1e7b04f9126d6818c9ce77",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/facts.json": "37c63d8a335aa822c0152d1edec345f2ba1cf0d8dc5deb6b2964570b40f6f4ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-btl4-35b-q4-2026-10-21T03-05-56-471Z-j3i3/codex-inline-judge.json": "31d729c628d9869e2299088002d85d49f43ed33aa47b1090b3d7fc05f6077801",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/result.json": "74c2e2ce46e29427d1eb44e7b660f344099234d0fba0ca35597b1ec2c1ce0532",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/score.json": "a12b4f97608fab9a101e99f6afae592397221369fcc1e7854719ea71aeb0d86e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/metrics.json": "975f794fd372adeba9938d2da0dfb4b055aa889334c85873a077e148aec2fa3a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/facts.json": "a80f090a4082343858cfb5b90f5909fda1519ab5dbbcf9563fdfe456b4cb624b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-btl4-35b-q4-2026-10-21T03-16-33-473Z-9m6l/codex-inline-judge.json": "acf80236622c4bebbb7022e5dd8617f1aea4c73d57a8b51e8217d92db35953c5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/result.json": "f1dd02b0148e75f373d659e47c24a66711f44ded08f1e83f88071c9f851570c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/score.json": "9108ed125b03bacffc1374873273788587406de699a32ac2a5be5e16fe9c3131",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/metrics.json": "6018c00f992bb4156960b34dc837bbee00e9beffc6c02e1dbf8178784987348f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/facts.json": "ed57a2e8869a17db0fc07cddd94ca63c738cee0dac463f00395c63a436dc2ddf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-btl4-35b-q4-2026-10-21T03-26-40-474Z-epho/codex-inline-judge.json": "141fd265b155919288425ff5ec1f100a37043aca6188c43c5a744c9f81fdb1a1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/result.json": "2a1d67a3e775b90da001b9b664b998a553cd6508e4cf830429169b9f8d51207e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/score.json": "04564141f0bbd3f30b0eff16027d350c590f77e31e99951a1f82c4018356d64b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/metrics.json": "dac164c5f8638477580eddffd1c5d42b9fd64464fd41c5fbf70062c4a0ce63c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/facts.json": "d49407772c13638bdc9d941a245b7a07877969977de48ab738d8d2eb2a77225a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-btl4-35b-q4-2026-10-21T03-34-27-945Z-kib5/codex-inline-judge.json": "1248e55d50303f7a7245e3bd9ddb6409daf3cab1da36c72adb72f52acba6e2c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/result.json": "1ddff855ab7a201d9ffe6c6d80ea11afa535eae40ebed2c9b07b0ae6331154d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/score.json": "6c1d0ce08ea2b4cd7874a3e5df17b707b586b3bb076ff2506dc252276bad6181",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/metrics.json": "2e6efab99844541803dd3283107338685a0326c880fd8a175180c92df85a247e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/facts.json": "a1aa8a4351fc2dcbd373a94049cd2c283186b5e090f2beb1062435f65231f642",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-btl4-35b-q4-2026-10-21T03-39-04-534Z-j4e4/codex-inline-judge.json": "d931e4803cbf66031be728ff57c6d4a05e81efc1d03bf41ce98e57df5581a569",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/result.json": "c2e2e3fb0fd2f121b6d57e723e6d3326c686076141496882ade137a69431e959",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/score.json": "99fbacd9131834aecfb6cab8c55ca6fc61441298a007baeea02215c9b19346aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/metrics.json": "213fe00921e44f79aa28434763c48f9994394f627557b4797cf9226f2dedd8d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/facts.json": "47264e0d85637583080f49844611a64cd36cfdf68487c4552a1500091ec1878a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-interface-contract/interface-contract/interface-contract-btl4-35b-q4-2026-10-21T03-43-28-810Z-4ssr/codex-inline-judge.json": "c5a30552201767d229644598ff7cb9557bf587b548b87794b47c6160fbe1e622",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/result.json": "f392182054bd8d9487702913fea0eabd6f15348495b17953fdd674bcdde845b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/score.json": "3834d6811275633655e35dbebbbef113ebd2299d4654d0a8d85fc4873337af68",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/metrics.json": "6c753031d8e54fd2601699eb3bc3b7e7248bfe0a25d398c13ca4b26cc3cede8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/facts.json": "f1e689f3ee6c02f50362fc8562010b65bc363718609da3e9980c0847d5727316",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-btl4-35b-q4-2026-10-21T03-48-03-012Z-58lk/codex-inline-judge.json": "b52ebbe87eaad930ad652d612e97afba296c4e6c69ee96626fe39ac7f6228554",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/result.json": "a247f180f5075359a7767276085a8aaa7a3bf8f478fee715e0161c9d35e79048",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/score.json": "ac5fd5e5bae4b63bd4d74f8758fdeb1486564f038e704b13b8129ad7d1c41bd7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/metrics.json": "1d4744e63eda580bf4e81871f6e71690a48eb9bc49cbd8ece016f435983bb0ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/facts.json": "c71088e25e2439f1ca86d51e2565f514eb13a47c13d00a0ead5a201a4edeb776",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-btl4-35b-q4-2026-10-21T03-53-18-972Z-mnfl/codex-inline-judge.json": "ffc5c4afd1fd5cbc8d8851f5f347a9fbffce628f70e2264c796b563d01e474ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/result.json": "e1caea5e72475a407682011baa28e61f72dadd7a44f63e15ffe15a9c47ed6707",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/score.json": "ef768613d670a4abc996b4b9be4640d4689a5035580608c98d7b584004f867da",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/metrics.json": "f74350a7d5edb2a8016b96ea96c4430da1e4fac10c923f253512fcccd38bd35d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/facts.json": "a0e7818566646ddddc5d952553d41f861a38d0f6a633572cb54455cb3f4c2673",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-btl4-35b-q4-2026-10-21T04-03-32-360Z-l0ev/codex-inline-judge.json": "366cf16b50cfc00e6cca3f4f6f7ad071f4f187d96f13de34128af48a8abb914e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/result.json": "3e1e12b27bd0a29a2fb523429fc575d366017c442d348fc2e1c450556e660969",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/score.json": "b7009dbf38d432a188d649364d95916a817ce4711ddabccff552390011370823",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/metrics.json": "e237cb9dfbbf7d70f416c3db49641b2fdefdde3ae367b3fe23de98ae2f8c289d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/facts.json": "015e7f81ade2762e09e981ac456de8ed162325f7a6fc476e50c4c7690be783a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-large-pr-review/large-pr-review/large-pr-review-btl4-35b-q4-2026-10-21T04-08-48-627Z-nwlg/codex-inline-judge.json": "4978ba196945adc0a049be0207b4cfcd57c8b2cb9664dbd82a6d77d31457ebca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/result.json": "1b75980ca9ff08fbd02685d79f4c02a8d621850446a2f1abc2a0681cf6d86c33",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/score.json": "d58dcf6f20b8d76ccd3422dbdc2297294c1354287e202cf336d4b2ac1024fece",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/metrics.json": "3c45be0d6518aac762d6a996e4f3c263d8b036a1e0a3781c370168c67caa39e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/facts.json": "6385ce427e90758f5008904014b74f947495e5461e3cec4a55df707b8c5ccdba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-btl4-35b-q4-2026-10-21T04-49-25-599Z-9a5m/codex-inline-judge.json": "af7d8fce90f3519cee9bcf4dbc63cc6d3abb34fe8390b06a2ac0e2f7eb0106b3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/result.json": "7a89a4b0d2840b759245151f33f7d5eb412bd39c3daa98016349eeb299bc1449",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/score.json": "fb196df191dfdf307a088442c8086a1589d2da1177c606c08a30e11e8ad8de4d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/metrics.json": "6ae9acc5ae81ac72bc375b87519b5b970d47d86e47159bf433a8fea8c8966e10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/facts.json": "a102566e213430bbf87f93ab937d8f6d95bd606350ac91492178b778d1bd9aed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-btl4-35b-q4-2026-10-21T05-11-38-324Z-f267/codex-inline-judge.json": "601629775fee48eb89980687575d0d87a65caca9e30ea01894424c13216f076e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/result.json": "0b38b63a0cec08f19f5144f21732c2217cfdb08204de8c28762a74dcd3da4c34",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/score.json": "337c1d8e6f7759c849d5fff0bd6b6e16f557595af14d3b2d103c69d08eefe6c4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/metrics.json": "df46d3dd08b7cb525012fab83c63e14e7f6b4fe819ee574c802c0670058d50d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/facts.json": "16de5f9c1602eda37f97e99158425280a1fc4ec72f6d7c3cb5df6d63e45a281a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-btl4-35b-q4-2026-10-21T05-30-10-596Z-82ps/codex-inline-judge.json": "25409aa3f122b342c280f61a547644053bbfed9445be872983531a5bec28c231",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/result.json": "65f65759dac59925c909175fe70ba04c62ceebfd27eab421cdf42bfb7ebb9c78",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/score.json": "f5d6902dfaaeb0843d07efedbccd97f3a2614a175b0b9c32c37dd5716b7fbedf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/metrics.json": "7b7c9ee705699706e145827ff7eab1f2474aa21e8d658ddbf6ff0be9c952fff0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/facts.json": "1134a78bfe7de67505960ad75eae005b675f7e38b72273572c7c56b6bd228445",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-btl4-35b-q4-2026-10-21T05-37-22-721Z-3uwq/codex-inline-judge.json": "b84fe411a67c1906f7adf799771581e05920f59bffe773797819c228116e88e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/result.json": "b525be23d56a9101d2ecfa8d401e3d51c76f10e97cc046cccf9ca350b176d507",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/score.json": "a211c158285c7eaf5ba1122f271f99804d4c1e5e820a374f54f19da382033844",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/metrics.json": "a77a183f82a558ef308e8573ae7f94651ff74205a0a7e174ae150c7e58d16856",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/facts.json": "d52bdc1961260c32bdad120c2ce63eb908c5baa463dbf85b88a16e0ab8814dd3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-btl4-35b-q4-2026-10-21T05-41-45-310Z-n43n/codex-inline-judge.json": "e9ca0b6aeeeb927e5442508181960e026128070229b4d321a1d9d848782360c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/result.json": "4abf3309061d251fd2720b4a5acbea1b57e14f7f3948ae9c4709c0d8f18df982",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/score.json": "d8b99ac2fd2b6bedbaa0c5b6d0227a06361173b2f443190ac6c00774645cea87",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/metrics.json": "e8e9033707767594554531e8d0efcba15e7f8c4cb16996341a10e4ab4b2c0147",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/facts.json": "42fcaa4e3b4659985d5b1697b7c8db874a101f2304d07582f336ce753407d29f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-btl4-35b-q4-2026-10-21T06-02-42-288Z-gxyd/codex-inline-judge.json": "76b345a9f08930cc04f9c96b1a6f5e6d6f6b4fde68979123b8a4e59a4b870caa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/result.json": "7cedb3a4a5207f393cc0b591cdc8cf3e6e44294d0f0e4751f2de311be0217df8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/score.json": "3e594b4317b06c065c42be300021fe6fd2fbc7d1a26836633ff5dc248f29c49b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/metrics.json": "5299a8abc585d405e86ca7b8a3286743c20c28b03429fba21e5ae33a86c520af",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/facts.json": "fc670bae17970e1cc355cfb70b1fec5151f6e1e351f14a003f182e24ea5200c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-btl4-35b-q4-2026-10-21T06-12-20-357Z-59f0/codex-inline-judge.json": "6e4594b2d1dbdb07fad10533f23cd698ce609f14dbbdbd11800190d751e5809c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/result.json": "e069cd46fd01a2748eaea67dfeea1a7b4b5b65ab1cc13f6800c9224dcfc0a2aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/score.json": "7015dd0eaf5b60ec47d3cf3b948d25d3dd80984af9bb370cb0ef7c50b88419db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/metrics.json": "bbba34355aeccf7c079b0ed590278235043b06e876b0c4e32926c3a5953dc8eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/facts.json": "d18152d4a4c545cb402f3c2fea56c2a72c22addce1a53d96c70d8b17ba01ddb5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-btl4-35b-q4-2026-10-21T06-56-10-424Z-upq2/codex-inline-judge.json": "1f904bcd656ba10e1eccb66ec3c9cb10ccd8e2fe8808b12d345601ddc7155fe6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/result.json": "15789b03a8260bdd99b342aea6571831956535bf261f8586ea9b49232b8d84ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/score.json": "20a0ecdedc5d88a2345380dd4262423a8ed8fdd27e782221af77fde99e0704b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/metrics.json": "fb8bc3a6b6ebbd0d6f1f1285a8180dd3f41a743a24610bcf1bef6eb87cacad0f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/facts.json": "21935c3feb08cf9044367006a6f5ef07d04d5adddd08c8cb6f2d0cce46eb5824",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-btl4-35b-q4-2026-10-21T07-03-38-329Z-4e5o/codex-inline-judge.json": "ed659fd56de191f6837e1f586ecb8f5c0bbea2aa58e2f4211c1d3ef9ad9fe770",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/result.json": "c152337c2a9810dae2e80ed1730170b255029c30d933b0e6549e4fc5e11f6e87",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/score.json": "0b81e457c144018a2e9bc1017189310d275bb1cad27b92af878ccb8ead839044",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/metrics.json": "0539ff3025840aa41504eb2ba2e1787528a179d838752b14b51f43f4734b111f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/facts.json": "8db93b454e7d8004b71044822379bba4032c6f32a13bcfda9243fe319066397f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-btl4-35b-q4-2026-10-21T07-16-03-166Z-54jj/codex-inline-judge.json": "3d7647109ba5613ead709f0962cf7fd026c7a2ee9fad816e8c3cc34628b9946b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/result.json": "a6298f40a2d2b7d9aef1a2ec66b16e8baf7e9124e4e6cd845b537281543618d8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/score.json": "d5a313dac041df3b19d6c8637ec1048a19387c37aaf3a7c84e8c2a214bbb125f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/metrics.json": "2ab99b0f4a58c87871bd6c56a490215475ee65e1970952ad06d70116e4474afc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/facts.json": "1e8009ca9bd10e87dd4ccafd7d8974c76a37e1ef6b2b5efbfccdad3ccdd36eb0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-btl4-35b-q4-2026-10-21T07-46-21-341Z-368w/codex-inline-judge.json": "861c9eca789f910746f5b3fea2f0e1141c3e30be702bd5353e185516db3e8ad8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/result.json": "0a1840623f82318a552071fc596fecae9f2f39b8030e9fea18cc528364e8ac80",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/score.json": "8e1a79ea8bf33caac205dab178e12249f3de692446ffa17b8399ba70c6e609b5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/metrics.json": "1952c07a4c95abed43f97117fc8710481c24a8b8552b8b0ffc267c4da8b06b4a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/facts.json": "5463977201132618fd63e784056454c7e6d412d8afb74b3d6acc0a1d363a6e00",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/btl4-35b-q4/r3-btl4-35b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-btl4-35b-q4-2026-10-21T08-48-54-736Z-8tgh/codex-inline-judge.json": "a3b33153a8728d9e67a7b34be10a560b7cc0ccdc0f3857a04e4e04d927cc5f18",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/result.json": "2da374c5d9d437d688f7b0e91d4056bfea20344a4bc2e40242076adc18544813",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/score.json": "649d1aeec02cede00c9e8711d92608b34da24d5c789f666f12dc51ea78a694a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/metrics.json": "b3134be9727da38d5f78920654afc23d61f4c55e2b72e504ce8f00174a54183c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/facts.json": "b7ff5e68ffc1a3eb39b7467187aa255e48e7960d0e9cc1d6dea0782fd952d408",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T09-53-34-651Z-c30w/codex-inline-judge.json": "867acc73c39fc4f163486718b9747a81593924ef8b5ac7508f6eeb326c6c6be1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/result.json": "8eeedf43137c2f6dd6db8a869347e7d236fd81dc9b724261ddb404a95c70bc03",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/score.json": "8289189b7e6dd793ab88edbee27b5f7ce74577bd6450b6c100288a6252d91024",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/metrics.json": "6f6d0abd24e232f7fc80f8e98714984a4a6e03055c3dcf8a7f63a46ab1d8bdd3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/facts.json": "7c084e50a7a33e8b663345ca512f081c73026418c0fbe0a1f165e63214301ea0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T09-58-42-120Z-qar4/codex-inline-judge.json": "45cb4d780b0df9f5c755d57335d33a1f43223f123144267c873b7abe28097aa8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/result.json": "18e6e2f1682544c46bee1b1209f25b6b7d877e1cde5449fafcfc05fb1016a09f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/score.json": "29dd7e7bd4babbc8ea66bfe09cc705582a47d90f229b839588a67fad97de77c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/metrics.json": "cf4fa440bb58f03de3be455edcf79565408e0152387b81e159a03fa9fd71db12",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/facts.json": "e7ab961123f3958586d8a4330d0b281369d133fa9758dcb7ac4d7c5048cb4cb4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T10-05-21-340Z-52gb/codex-inline-judge.json": "fa09a0fe12a7006f4daff33755a40586c733c4ac70bf52beb22fb91b4a6c22c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/result.json": "a6c78c81797e8fbfcc7bf44fcd9a61914a11bea656dd19f5b27bbc73c183a48c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/score.json": "09e957656205d1015bd881f0c97e96b1e4c082f1b0a2dd9a0beb1ff5acdf2fd3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/metrics.json": "fa207028cb086d736d04858e550bf63dbabc6d2cc5b215108565c579e2ab3e9d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/facts.json": "ec9be1884cda147264cf0855ee2678827398688d62c04959a88c99042730e606",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T10-10-45-269Z-p901/codex-inline-judge.json": "b65d9d61c7839dc5902609205114a3dc7d411d963dba61bbfc56e6eeb5d466cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/result.json": "ea6b2d1deed876d5b086e17ebcfdfe3bf271a95d2001e6e3ddd670629300e2c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/score.json": "15eba12beefa507f605ed836f55b70af206e03fe64ffe19e9b5e33d8fbc600b0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/metrics.json": "76ea4a8d2eba5303dd0d9dbe125aa851ccc3f481901217b36fe4561e0b12eaf9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/facts.json": "d835aeab6ea90e8a92220a92afb5dece32ffcd310e1f077939874cbec54d2484",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T10-16-21-248Z-uoo8/codex-inline-judge.json": "023cc6c2de4ccb4f83e15036a3f9ae4a00ceef5e5f7e4ac2d146239bff9bf60a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/result.json": "5942e82ae7afcbcfb1be172b65454264293f9cabbdd08d8f40d679e25218b66f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/score.json": "ff59a40b1c791e3e06b60c11612ca82a743a4ce3712ec31841af74c05a2938a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/metrics.json": "80b0c961e18ec1616819143d85a46bcdc52256669d314e1cb9b3f707d3cb1844",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/facts.json": "46cfd60d88d43a9f5bbfe613f428dc8afd4f80725159a6e0f6ea4b1c42062b02",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T10-21-48-204Z-nvjr/codex-inline-judge.json": "97c7106877af8bf0d3102e33826f8d53a2dd32076a68f87c6a5bedc4efff2806",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/result.json": "8dbb0d17c84cfa349cd005514ab091f2a40b540d437eaeafc843afd13704f919",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/score.json": "8e347be0ee0759b122e87d663649db838f3050bc95485ba219401927063e3a8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/metrics.json": "ca6cd571f7af2a9782963815ec9cdd1f86363d4f6c04b7c07e2a1d9f9e8fff7e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/facts.json": "f586ede2e7756ae857beabe7d7b05578d6ac550086f4765e3cbee5a8ab9c7fc1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T10-26-37-717Z-8zrv/codex-inline-judge.json": "bddc88d1847682df90d183488722a99a94887527c6e504b8ee1adc8a63999211",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/result.json": "df10874de371114be611c84dabb4532a0a68e5c6732ebc258d5404de4db2a8cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/score.json": "4a1a9c41b2dac5fb74fda3892d2ac909f45d83080b78d78083d6c1a477ab0215",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/metrics.json": "9c904ac6443b3c9a8ce48d26274dc175139daecdbae207245f4e5f0405194fff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/facts.json": "fe00d78035f750b461b6f1a97fdbd17852c5dafd3f603d5011270c67af1edcbb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T10-31-41-802Z-3n4x/codex-inline-judge.json": "45e300c1110983b745f1f8d2a89ec8d088678591151a92ed6a24c64f1872901d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/result.json": "ea4a496512e876281a7025336353ebc62058e8e2ce2a201760436c932abb9979",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/score.json": "35f7e3990d99d82ca2f8ca14f5fa5558dd7e1c6afb912ea3894f2354f2a3db7f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/metrics.json": "6ba6e5532d2621bb91123484a81fbe518ab0fc6d721808df60c9a35f2646117e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/facts.json": "22bef327eeda6d795af21c3e62bac89c7699f4550501998dfe619d2cdaa8e410",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T10-37-36-368Z-r6qo/codex-inline-judge.json": "6f009852cfb41fdaf85897f83b8bf48a23c45d48100aa9ae9c19e5b7af0332ca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/result.json": "1a745eba2196746b8195915d71ab824c3a5958be456b451b4138fcaa4b7bb450",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/score.json": "810c98d8a97802f9ef8d0adfeff4ed08b4668ad18222c5dd1547955bd3c8c853",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/metrics.json": "604ed6265dc59f9704767bc2e99cd69c88608cbecc7585fceece8d92716b7d74",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/facts.json": "890458ca6dcb35313fc394c21466aa2d09cdee31961a05eede31a7efae0c003e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T10-42-10-746Z-97dg/codex-inline-judge.json": "f848915e7d5a108f453b53d1d89ffb9aa9036eb66d9d0d9955bcd880f019f64b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/result.json": "1a0de465e33dd56266b16ba3c93653a335bda5f1c16827c13a5149391a4d7aba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/score.json": "a28cbcd7bdcdf309837969f160fbd490c1c189c0ab3a0b4be6e8ee09d1a10812",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/metrics.json": "e8b07859ae71f794871252462d26c4fcbba4963b558ab29e201b267f01b01a2c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/facts.json": "caf1eb2321dd688008937be3748b2c45a031a168ca798ad50b49e3f5b593307d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T10-46-39-157Z-exm9/codex-inline-judge.json": "3d14ad3e8234fd04c8aba8daab762edcb9018d9f22599e559c774565454d1963",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/result.json": "71e5f2f62e4768fde27b2a32192815c9496d15dd6f773dba8a42864cbe4717cb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/score.json": "0b5d0bc7f3065c420ba5109e61b14112995fa705ca9e5a5d119fa6f04188a05e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/metrics.json": "2fd69599f90c5e5b9e9d11e7562547d06d4e04294c821d1060814f60a0a42e94",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/facts.json": "e748e647e8edf7c4f4475887306e5edd5ec35ab5a6e9f33f209a426eb6830ab3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T10-52-18-794Z-xj6g/codex-inline-judge.json": "236dd79f593c2012935c00eeff435348abe3a197a7540566572129485a793e47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/result.json": "3eac6fbc1ca42095fb25026c44a24457761eb12af1953bbc89a845d81c42f015",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/score.json": "52fe049da39b758e859bb96161a0bfe566b681920e32cb0080a2c77544bed4c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/metrics.json": "bb82c03a25198b53c0b3761cfe1f19e47b81419d5037ecfe13773fc01a3fe663",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/facts.json": "5873b5b255c6c044fee60c36e4791ec1eefda4eed84c0bc4fb8d0d2ee5506891",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T10-56-47-023Z-p3vn/codex-inline-judge.json": "8b2837ee58d434ae01ab186385cee2596a4d0b92d7192780abfd47668d1982c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/result.json": "af82dffdffb7296d048d5d01ddf940b364bb587ac9bcbe5b444f902a98358f72",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/score.json": "cecc048773429c25289f920ca3a6470fccd22d08901d5a8906d17f2a116b0c33",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/metrics.json": "2ad9b0802d4d7d9e26d1db7d596cd6b166cc4f02d001209b4b6f9fa7c618d5b0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/facts.json": "dfdfe551bf4ebd648c15cea2e928c5fd30abf08849152a817ac3667ffc63594e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T11-06-27-562Z-3kme/codex-inline-judge.json": "c9bf112bad2171c44dbc736ef71737f5d9f5eda832ad86b606281c2350efbe05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/result.json": "fd4f65922b434cf54a5a3d2ea927c362d8139354f7a22cc6512d05525afc1f9d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/score.json": "855afb037b308e136050c77e9897b2b15fb961b5fede0580199142769f3352aa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/metrics.json": "858f2f08b0f8ebfacc5503649719b39e74d8707f39bff1c0fdcfac12c60dd837",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/facts.json": "57d2a61385a32b1cb9395a1b1558e8c1ec39ea0c2dc9e039578dede0dc0e0f18",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T11-11-00-918Z-p57y/codex-inline-judge.json": "b6454e547b37d243ab6d80cb114e156f376d701acdbf8f8bffd234f9aea90652",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/result.json": "a8985d2e2053a7326f497cda9592b359c5549a9d915066a097d06aa12aae0dde",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/score.json": "66afb11ca495ab8c2a61480beaaff3f90683ba599fba3c37a30e2ecd18e4adad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/metrics.json": "5960d98ffc27bf959086d0f385913b0272a481855a29c63482a85d4fec9f41e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/facts.json": "1e16347319266eae1196b1e9862eeaaa619d9e1923593c03d66f329e25672f1e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T11-16-55-708Z-22vz/codex-inline-judge.json": "e8c444b29d300322630a561de75944aba674388b1d7bca60fedf61c6f448e1b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/result.json": "a3bece607d4004f46f1139e613b4eb6712ac191aba81b21e1c29701e0a870c1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/score.json": "0de4f0b5a0ef1ff71e85f2607b7ce43fe7e9c84800b11c03d8e84feebb5a0bb1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/metrics.json": "705dbcbc420581e3b434d603069c172596d24770f3b8421ac60666b0d37e5207",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/facts.json": "a74a5e8ed736ec3061bce89c81e34a36401d3deba1359f03155722254bfce415",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T11-28-48-643Z-d493/codex-inline-judge.json": "4678b84950031b9d65e088fe22483c5b2685904efcf2de1c39c0a8855615a70d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/result.json": "c06cb33353c71b1e6b2cb36cd53b7a9a1fba0c81d0619b6b84dcf6977a4efc6c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/score.json": "b80ca5ac8a36597c7cd85218683577033929e4f61b2bb1c482e6e7c0edc09984",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/metrics.json": "6bfdd664d8900f61ba646ed8a4287bcce04709511edd78facf07bd5480fdce2f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/facts.json": "4721d8a29531c799d4d43623728cddb973690ad955e87713ff04502507fa44ca",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T11-35-57-371Z-9pj9/codex-inline-judge.json": "9e68a2c127645175cd65e225e4202b3f7b5dac2456349b1643e04954aef4be71",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/result.json": "b67efab0e7084c16578f99b146a1564d33696a18eca250ab1aea5e29b06ef0d1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/score.json": "e27a2f9762f6524f560ac673986ab0066de6e0360cfe1e0992bc42ee573d7f98",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/metrics.json": "53b143c7a8c27e7b81292f61260139e1ce320b36bf1658c916f309de7bc6db9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/facts.json": "cd9eb569e7b0951d064553b96f654ccecdff9c5321e47da01d3c66de5d2dcdcb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T11-40-30-846Z-yy3r/codex-inline-judge.json": "e66e7f6717c8f9b9bb1b616a82d8ff23fe6d7b02177fe140e59e4b1c1aaac2fe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/result.json": "c7723c02b3610badbc475a546eb81fa160528afee84979bb47ce28c7f0975dc8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/score.json": "211e85e21167f4b8cd2e3689c8a76069db068de5b3f58e94f3bf68097c3e86e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/metrics.json": "d4f168e015b57d81c57b78191eacac54ce9ec1ebb954b3d09fe8ee6aa29f5daa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/facts.json": "cbd3d70e797ce7e26d7865aeb9b5a19b4e5b100337feee34aff84742cab7fe17",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T11-46-51-144Z-dkmo/codex-inline-judge.json": "9a3c53929e294bbe1691bb7eae88ad6cfe830a7029e48319cff431d80bb39cf8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/result.json": "f50d6784b9093e43ca74a0819ae8f1197ddb29213d1c6f39b7152743678f5578",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/score.json": "3888dff6d2022baab2022ef564cd44ae5f2351cb7e2ca1c3f5d852e9817871ed",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/metrics.json": "f78bbf31ab234ffbf3cc752c073bacbd0cb4b028900a686e33cb12e190c399ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/facts.json": "e9c5697ef1a5e7f155986d052d0f71a5447d62f3337857f98e1a4838395e2c12",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T11-52-34-757Z-t97z/codex-inline-judge.json": "c6235324cfef5a184178def621944dabe5a842f6338f6063d98a9cb793361d6e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/result.json": "d4785c837ccb966cba38c7d14697b48bc1ea22b40381dcbfcff3f966d152c4d0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/score.json": "f036ab4e0381a70405a3dfe2cb0ad220f55c283f44b64b2417700951b38523ae",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/metrics.json": "a10f9e2c5e309541b83372fb70391c21650fd408c8347f10a1338fd22558da57",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/facts.json": "7836e5a07a191055f3caa60f24d694397f76b7b6f910705c8e330faf2d43b733",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T12-20-16-906Z-7wqb/codex-inline-judge.json": "b7d7ec9241d0f3edbd7f56921b452085cbbfec221ad5baba86fe43c8af38b05b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/result.json": "2ec0b8bcd81da537f74ebdf5a78e7bf7dd400460c9253c9d47dd42cb737b32a5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/score.json": "34cea62b7635dfe94504cbf1377204fadb8c035369bba037c243fba60ffff593",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/metrics.json": "8a3a7eb50544193db0f9019073d101b6879543157f7a7c17d032fb668a87604a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/facts.json": "9082a4625384a0ff98ce282511e49ed8b7bd42b1d504d67205e5b18ac4e47345",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T12-36-26-067Z-sm0o/codex-inline-judge.json": "e681c11ec2f7697891669c6179069e021a05045cd7ae17823b1f71ac29b90442",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/result.json": "4e495dc2923505c1f3085a7f2cc2a8d2e43f852453891ad917d0ada3331e1721",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/score.json": "8227373f8e61ad6a2c8be910f6fd370bd3365042c1ba56602d5ff51419a90c27",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/metrics.json": "ef2f9b3c837351db4307774544c7995467111f7ea6d79bb82bb235dee5dc4c23",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/facts.json": "74fbe39069d9cae83974d13254fb92b9cdba918b026eab032e384082bfd42cc9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T12-44-02-747Z-stb1/codex-inline-judge.json": "7e9a2f7ce5d276b0162f0251b402e0621dfb8175bffa237a8516ac573eaa6600",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/result.json": "def6df30faa4880638ab142ec9dd10a3b6aabf9678165d3abd2e44a89adb57d8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/score.json": "571c3a62c4e18ca7c5edc0eeb4c55550548ecb5b28e1ce06f8daaf8ee9648c39",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/metrics.json": "c6cf5125f8cad9e6cfa659f5b23bb964807cb33c4853890c88347a1c866fb566",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/facts.json": "bd39164826101458dc0b90a66f9414a9e0b1130725ab2a84aa344701064b6251",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T12-50-26-410Z-obvl/codex-inline-judge.json": "e235bd66b3c17afaeb99020ca66368f2cf142bac7679e9b4744c274ff0412682",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/result.json": "9722c6ff5aafcde5d1bec3d5624282ce1457160bc78c3bb5c9eff7251796dc4d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/score.json": "773d06cef2dc4e0a57b6ff0c5b1adac66ac344ab1c4c64afea920145d7be7da4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/metrics.json": "3f9fb6357193757f9cb43d18a851e45f916b7d5e82d27abb6d65c31ad4e0cb69",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/facts.json": "61fab9e8584820eb684e96efaf38dfabd60451685b997addb2878c5ff220c559",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T12-55-21-219Z-j4y0/codex-inline-judge.json": "fdd181f9957628b445246d9771046524f70e0bd3b99dc5ce69665ce4c9452a45",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/result.json": "8da554e554450badc8090335811e104b847c27fcd503e3fe6ddf4406e06441cc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/score.json": "04b6554acf7fae2c48a76e836e5b4bf36161501d9a0ac5cb4692e9b024f305b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/metrics.json": "2ce4a783aab60ce3d59987d4557bacfd48fffb96efe874682c0ab1702b8e4994",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/facts.json": "beb1f8c1aa2b346945d4e33a1ab7c9802d03a49370c86781ce3d910ff39f1e8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-21T12-59-56-472Z-6i85/codex-inline-judge.json": "a8ced79fd22505cd43decab9f5878ee7726b1020cfcf61408750a1cf3c981170",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/result.json": "d4efbf818ed25c393d05605a90a995bd2d252410624582744f9bd14945de0d87",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/score.json": "f3439d3e231d460bf9217ff8a5b8531f914c97ae30295c915316b26527b274e7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/metrics.json": "37c1b9d18c4ccadd1b4cd784fb9d004351743314f98d22d8198082d40a75ae98",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/facts.json": "c9b79552ece7d4a800ae557dade49a28ed22372cc16a927f5d4605eaee61a786",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-21T13-09-58-116Z-5m8z/codex-inline-judge.json": "7a0ffd8cff838c53aace4acef0c64bcdde9177c2e56f363900331b8bd96921b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/result.json": "ff290d82733ac6f19e1aa1ee6acf6affa4cf0112bb684952e8fe28e22158b609",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/score.json": "e55ff67f7490149cf525fc44f734468b6b0fd68c4c92606f5f8f0d8836dbc980",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/metrics.json": "b370ab7f7c35acb46e3eea23d9a863c207821acca1d07c5331f2c4ede15e4854",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/facts.json": "d2021c17a9ceba1054d6c85340b28d1616d44216217444a1cd872629c6ef1ba4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-21T13-16-09-239Z-s4v0/codex-inline-judge.json": "19f5807aa248780d74671ca646f7cd21348245021574e787a9ea98a2b680d689",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/result.json": "a18497a9be6477174ea785e9b4ac92e070a27838696dd2b29a37f32ed28706bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/score.json": "185bd652764b3c6caf64ae97e76623b8883b68c658d60b32658472806a5f1005",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/metrics.json": "30313b1b6ea3d80ce906883dd29c35a315f63d4ca4d93f9ae1f0a7c44c4e9c93",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/facts.json": "17069c0ad245d4b6866aa90e3082cb4a690bd3eef4da3a4d31b8126feee08d8e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-21T13-24-40-873Z-fedw/codex-inline-judge.json": "c2b44b0049e3cbfeedee747d121910f8ae4624a83809ac39a8caaa1982ec6328",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/result.json": "5d56f565b68156c20051a6969cc1f66276945e073dfe482fb81474d95ec85257",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/score.json": "15ba4df0327f18f2634f08e66f28f80506d65b873db5ab176459d99877e46259",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/metrics.json": "fe87d1e0d267a06db94e9e0ba67666c6025ff2042e3466c8bef6d91a563ff567",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/facts.json": "3e4c562c4d7fc42d4e757bd6c45927f0e95a21d26ba9b6dea7938c6b2a4288f7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-21T13-30-42-019Z-pq56/codex-inline-judge.json": "e80a9361674ff6d7b2e9700db8c979ba0903ee5b52e6d837a1d8057df7f7682a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/result.json": "08bc1e3b263e1d9503e42acc317f6bbfd73acc1b60cec6b295f4c8a6ec4758a9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/score.json": "3d9f2a2937c1259abd1417858aeaba7b0d65ed7ff4ca4f630795b79a2156ed5f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/metrics.json": "afb0a3972dfb228814cfb2c98628800e3e14ddcfec069a3fafe31229a59f91e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/facts.json": "e306c159d77f7710790973ec4491ce2ac279a852fc8494ee7a178dee4d21e06a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-21T13-38-35-294Z-t3ba/codex-inline-judge.json": "b70a0d78e9dc1464bdf7cb2e176a1726bf5c272a12a92e3a9ffaff061a325120",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/result.json": "6a4f94aaeb43d9d25f890b2adcc155d8c88210ea14240ee242d73b57c361f425",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/score.json": "c03bd944bd6253ecb4b70149b94a7446740b7b1903d1ee598c661362aff2aa68",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/metrics.json": "b6dfc649151e73f2bd19370d275d409c1cd8f4922d6ead1ab335c4518137c683",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/facts.json": "0f589f9084b34b47f23e7df0854e36ad10353ed96fd46f0f26dce21c76c2ba85",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-21T13-45-07-702Z-sbab/codex-inline-judge.json": "eff1fa9545e629364a7535bb4a1cafef9482d8f84a6fc18afb6d2b5170e16d24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/result.json": "cde9ceac147c861b9574b6ff46cac6d7f6a8dbba0ccba6d85197b13a906c6563",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/score.json": "b43e41340e14222ea5b38e81528b77e471c3cad81ab1ccdec5413ccd22fbdbad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/metrics.json": "3e7b2140a1a0ca318a9cbec3dd4bf56c912c5adde81c7b00ca7ae7315455a5dc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/facts.json": "8eba82d45d6cb731daf8a81461c34b76f398555e0e36564f3856b49c62a36990",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-21T14-06-11-934Z-rwnk/codex-inline-judge.json": "b477e74b1f96abbe4ef992fcb12ded97e984773a12a62c28773988332992bbd2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/result.json": "e9ca129862014c0be699f8a679353245218a19385fadbefd9264c35cedf75b1e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/score.json": "0ee309fe8021e8a9534b085cf5dd5368dab9942c662ad3ea9d49f3308f117de7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/metrics.json": "6cbe20d8b565d644b5516da76cc0318e39be9f41a6d1d0c1017ee667ad738973",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/facts.json": "4e878581db451d13b1fb6a9cf7f720b09ccb1717fbbcd2adbcd6de6db45bdf35",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-21T14-30-43-336Z-lnv4/codex-inline-judge.json": "b9d2d6d6d81f9b38740381a6e363c865efa2c5f385c7b886774db28c6c12ca9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/result.json": "8e0eb78f41c42a1854c9f634d90a5e8df070966484adf8784f36373c3f94bc6f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/score.json": "29d490ecbc49ca40e2ff79d801df885224917271b022581f9aca207cd24c4dc1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/metrics.json": "fbd9aabe24fee4fa2bd672ad8c2925311f7e522efaa2edc177d3a61adfd05712",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/facts.json": "9e9345c6c248df32b6875b4ee8c2bc8bcba14a3b87a9a15ea8e1b72c6fbe9c1a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-21T14-35-18-056Z-kbkr/codex-inline-judge.json": "b9f0a3812f9aabc2f5e6da66afeaa18e5ab63732ce7659f4e7c620a256a60115",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/result.json": "7fa3a6f9c5a8d7bc5461b6cd446063ed26fc680607fa95283ce9b89b986fbdfd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/score.json": "0fd123924e79629162971aee841487cb1c953a16e6177afcb3bd69983430627b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/metrics.json": "1c498f6122442de08c275360e6b379bd28c3fb569803c843f23389a3099ddd02",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/facts.json": "6ad9a1352be9e4eb94c79646385b368a22fab45c4769ede9cccd27ca802d6277",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-21T14-56-44-140Z-so9p/codex-inline-judge.json": "9d27373fea1163d506b4c812ef9960821933563ebb1a4b7031e145304b2a6732",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/result.json": "faa8a567a64fb07f2f9e6f3e53fa4bba9598bf1751db5f98c159ae5d3cd46678",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/score.json": "444fb327fa5027b0c9d28a07ec49d2ad26257919c18cf6fa9840a8d548a436d9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/metrics.json": "9d169d64ac6f120fefb23a6197e67bee4a80e07c8f912a9e35b2586a4ba59756",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/facts.json": "a22a8b9778d22858b92d4dc3bc69e60e02a733c599a9998b7e5feba996d5d7c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-21T15-04-53-128Z-tt1h/codex-inline-judge.json": "2ce453a14360787075d63eec98d5c793c180c07f37dcce291e56a6586b17b66e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/result.json": "11dc47499de3f3efdf30e08f4dc0cbaa51475e3610135e1d0d0180dd99c726b8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/score.json": "912fb5b798fde442cd934d2d5c6563851e3a0be3eb987183ff3595fef5974c90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/metrics.json": "0cc5b455137c8f7d1dd1501b7abd6cb8821cafd3701a648be11e421eab9427f9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/facts.json": "1de6fb207d6d4d466d9d49b57168d72e906af81307e82063bd3b0da1179dc5f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-21T15-33-33-597Z-myn1/codex-inline-judge.json": "abb4b69c3782591a7873cc77fac6c9d38911dd4cabac048dabafbf015d73272c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/result.json": "725491e9600b5259e44c8da95f31d127ac88544337dbfc1825d3298ba245d5c5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/score.json": "28b018ae66cfe156fd39468710eadff105b0ec074a92cb658301f1efe98356d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/metrics.json": "1b38b45df8aa47ef085dc0539c30d93c9af04631a98dc9ec3efcba3615d75d2f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/facts.json": "2009f00a0e36b1ccaa48656d403f6e77c7c506bb7af397673f6b76414e4d29fc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-21T15-40-14-343Z-mcyc/codex-inline-judge.json": "a43c307aa133c03f84f230e7d57f59f1155ce2eb87836ad89066b6744f0496fe",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/result.json": "fb749d70d209d76f27a86c86eef93e4d2c7d08bae5b69e22467b0bdb54c7deda",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/score.json": "86aea5bd8097a372c61a1325cd1751f016825ba04f9562d01cdb85c6195005e5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/metrics.json": "babd408d1bb42fe5959a03528d4f133e5517f0bc6f30f0873172f944797a0088",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/facts.json": "c12f6ae261ba0884ad99acf6504290a8be59365c2af36601818e2030c6f50d27",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-21T16-18-23-023Z-kume/codex-inline-judge.json": "32bec9e2d4d15155da9467defdaa5b665302c87ee1881963d6adabad406d6e8c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/result.json": "c8c66580639073666c3e3af59931040f367aaacc64d6f0f302a4f25e2875ad1f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/score.json": "fe8067bfe24eeab5757604ada57e07c240fb7c59302b4b9f3cf894dabf223385",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/metrics.json": "4beb38d928e98bf41babfd8b4e19a6267a531b6749b78091ed538c45595d3231",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/facts.json": "7c4b42e935801d7cf98b64055bbebd7fe6bf5bab5084d38fad3fbcfea508e4f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-21T17-24-59-702Z-63t1/codex-inline-judge.json": "b2435ecebb03dea48ae3fe646fd6789ba043fe4e1ab55b0ff41da3c22ba47e1c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/result.json": "ca35831afe2ae08ad62284175f6d8062da38e4c0bb78c3eb1ed9d7defdc5498d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/score.json": "21c80fab4e8e22cdbe2172a5e6e3ac3542dcb7b9eccd0fd812b67373e0dbbe8b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/metrics.json": "11015bd61bb2544b22757df12c4a8964b955d425f2e1f51295a4142e609a65a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/facts.json": "7b24e7ac709a285ea4a48b5e4c972b265bc8b296a44a6e6f913e16d615c32de6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r1-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-fanout/craftbook-author-fanout/craftbook-author-fanout-ornith1-5-35b-a3b-q4-2026-10-21T18-47-26-960Z-kbkb/codex-inline-judge.json": "8bbd5cf274e4b28b5c2ef09316cf75a89f42daee64b2b95b96a715580b426e23",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/result.json": "f4ffa0be74936fa393bc7f505d899db88870c42270ae84d93a3803f36e88292e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/score.json": "ad62a5df8b594450d419c94ef2ff139474eb4b8a7acd4bfe02301ffdb43c5af7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/metrics.json": "8dd6b850e89d4628cf7de816b6f6e67b015148c55fe5a8724d3b03d446b28816",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/facts.json": "a9dfa482488db2582fb7e25941fe7590e49324aeae055272bf5bc16f72aecbb1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-35b-a3b-q4-2026-10-21T19-41-53-524Z-dx4f/codex-inline-judge.json": "9beff86be843c65f177ed413b9caa7c7b445ad96e2a50a4a328b5d23dea49d0f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/result.json": "be49105532adca5acc613606dec8788bfa103450514613102cdbbb1aa6086b65",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/score.json": "82ddf67878c7b991843ec0ef7826a73c82743f40e10134ee77e7b11a7d002c20",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/metrics.json": "8e62705b5fdff2825f17b9727062958eb9ff939b3ae453e3f5d1dec8cd38e95a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/facts.json": "7034782857a57b1bc98e95174290302b6b10a710035421bc2c351e190b1c330f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-petshop/petshop/petshop-ornith1-5-35b-a3b-q4-2026-10-21T19-47-03-112Z-n5ex/codex-inline-judge.json": "ca7e8e9dda83e22f296255f6617404a82fa5e83c124b46921338c252570c4f24",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/result.json": "ef8ffd669e66c22b74fe87347ff0a4a283a45b9507a0917cec8cc4ea12d5bb9c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/score.json": "16c8026e7abd5c93f1105bbd043f67dd123b37ec825ced91b2e055ca2e35dee4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/metrics.json": "ad171295a057a3420441a1246202698a806aef363d6e4e4edf95334f0decebcc",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/facts.json": "74035f30233197fe808c8753f078c11abeb095322b28f1138be00c6ef4f9d4b9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-35b-a3b-q4-2026-10-21T19-53-07-040Z-qdwn/codex-inline-judge.json": "05ebabef7fdae98b8e9ed91850501ebfdf8c0521a44f97ce6e1bd91b68ae0f02",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/result.json": "f6b260492e9d4dc72603ff8d56faae2e4d45ac0fddbd1834d90306bbebc4b3d7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/score.json": "6eae30d31b537e9ebad3c64318f73c4ce3997e7afe0d7b758eddbbf6fb7bf2db",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/metrics.json": "6eab56bb91506f8432ef47712b6fa23b99c6610cdfe46dfdc01a71009eb8d5b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/facts.json": "0049fd2851dd742cd8ca93bbeba9c530d8eb246b8f99690210ac2629e49bc5c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-35b-a3b-q4-2026-10-21T19-58-30-243Z-132s/codex-inline-judge.json": "18feb981fcb43a129b6edaa92e0d4f555129ba92a643b5ac1f71b7b3e8aded28",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/result.json": "d46f287f14bfa85a205679a49e3310f135b007f9d4695f89b8385363173a5adb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/score.json": "d49ade82c7efa50ecc02b35005253a38f6da7884cc2ee54cbc6b0930b4fd115a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/metrics.json": "e197822261559d3828d8747e07885aa828753a59dcd697dd752e3abc899e021c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/facts.json": "50899bc22625ea54614cba1f914585fa306e223098bd86d4f69dcb6d8607b5b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-35b-a3b-q4-2026-10-21T20-04-18-037Z-1jfr/codex-inline-judge.json": "12aee0984614d8564889aeee927ee4b52985d4d94469d3bdd2664580b7e2a3c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/result.json": "52dba2bde6daed329f970a9761d56f4b81331ce23fc43824ea41e5a2ac6c9c90",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/score.json": "66ead81bd518bed9fc4223829134980bdebf6990d9d8b707420133fb1a1937bb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/metrics.json": "2a67bd98d0a3b8141a2e18113cb564aaed6be42e600e680f029ecd71ee18154a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/facts.json": "3b3de16d7f883ac626745299ad340b4316fcd3717491ff81c0e1623c27fd6396",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-35b-a3b-q4-2026-10-21T20-12-56-468Z-uyar/codex-inline-judge.json": "ca3eb321fd8d6eaddb46506357e87100280c70e318c91018450613cade3ded47",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/result.json": "03d3476fefcc84d680a36ad3388bd6cdfdd51fcfcc9d14161da45f9e5f0f439b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/score.json": "509232985690bf5884fa0eb2da3f4de62598407815174b973b8d7c6d95e49d61",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/metrics.json": "352cd618a5cd51d1cfa0d709deda699457de44395fa0c9634462c8af860fa7ff",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/facts.json": "3f42a8afc8701aa95173c9b8f8f31d68497916a2fedcf9329df41f41851ce1dd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-35b-a3b-q4-2026-10-21T20-17-51-556Z-k0b5/codex-inline-judge.json": "899c51b8ed67694d8e548823eb843ccafcacab7699c95f7ae7cf3db7036cdeb7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/result.json": "428c48285997a61650a6a868d6826ce32c1db44291618e561c18433c2b161018",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/score.json": "585cafac9fedeb49cc18330f9e976a13a58c6a8cb0b667c797272122001a3908",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/metrics.json": "b9d3a14b9691aa198c3f1d0fed07152ccdfa41cfb4adfd1e95737dbae446fb4b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/facts.json": "5dbc3f6025c7957ea57baaff6d76714a4788c480c8c1d598230cdb5f780bcf99",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-35b-a3b-q4-2026-10-21T20-22-41-378Z-le8n/codex-inline-judge.json": "fd96e598ebfa5c6f85b53e8e9473e0491d1a05aa5c29895e53290734b7dee8ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/result.json": "90d85e36f17cfce62327381242d2aed27cd0616ad6a3d49cc947c8b43ffb35c2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/score.json": "3f487c9f8d6a8c695a08edb770625010572d33c5ebebbfa7e70033c37d844336",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/metrics.json": "22027b80bbe307f916cb6e4e5b8af9e909c1c31b50b9364563d55c0587ef8079",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/facts.json": "97581e0cda5526cdbcef66a09a22ec332a1e722af9033895e207f1cba037715e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-35b-a3b-q4-2026-10-21T20-27-52-771Z-unw9/codex-inline-judge.json": "61660d0e2af301fc6025d671326cf824e0b8bc0dadaa153acff53c1cf1546b94",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/result.json": "0a969766df95cba8612efe9fd126162eb49432d6aa1611bd5027af65794650c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/score.json": "183887044d9e212d2b5883aeb8f2ea4808edbd752464aa0fed0d8458eb58c4a2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/metrics.json": "03f4e38d991a61f5e749ce448afcd6dc9ff6a7a54381c7a7f5c4ed3908e47754",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/facts.json": "11db0850a9305334b2716374ddff7571f7524e84036a939a6613f373dcb19ddb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-32-50-178Z-0jmj/codex-inline-judge.json": "f89b6c256a02af6db85ff0328343551da4454f56b828bbad7c7d75bc2cdb6741",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/result.json": "23b47bf47e063e366a3bc64ce9f56406ba83ed2c8d32b8eadfa1bde1dc6429eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/score.json": "34c3ddf69e8855f5d2f3221dfce43c7b764ab8afa6b8be7757e4706420c6ffb9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/metrics.json": "4a1513fe650953df15d371d33bad084b72b99ca2b844a44fc606b98627a747f2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/facts.json": "2550be87081ebfa23d56820fcd78ceffa45a1f7b43d838c91ab614efe4ddf301",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T20-37-27-733Z-0mg2/codex-inline-judge.json": "69625d0d16f45986dd8b65a58ea8c440b35527f0145b7bffa988e0d0f128e749",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/result.json": "bebdbb1d7dbd59aea00a4387ae0616839a83743773132faf326c72a27ad5ae1d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/score.json": "ea70009561ee20b96267dc3e4a6575b18f569b07e77664524f59395cbb8887ba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/metrics.json": "7d3e5d32ff56b2d320ea23cad975d72999d74cd588a1413316b174c8fe8a386a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/facts.json": "871ee0e08c05bd215811db9bc8cba8c54d5505ff0f021c0a4ed714c035493948",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-constrained-comms/constrained-comms/constrained-comms-ornith1-5-35b-a3b-q4-2026-10-21T20-43-09-028Z-tmwl/codex-inline-judge.json": "a6d1e1ba2e00b55c3638f8352c38c3303ee5617ab8fee2771f94d85a058f846c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/result.json": "a7ee93b49449a551c3fd2544ffb9653bc8238a75408370aa02daa1f7e994b4bd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/score.json": "64f35182b585adea5052c42ebfa722b87c45c29b0fdbd483d066536137419a8a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/metrics.json": "51a48059c0f8d0168b1c290277ce2f940aeaccf9e9e649b64d332d854caa9d1b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/facts.json": "b10eb3cd8a5e75c841547392b6e3005f9cc64896588eabc6c7cd271130a35de5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-records-intake/records-intake/records-intake-ornith1-5-35b-a3b-q4-2026-10-21T20-47-38-793Z-3j0p/codex-inline-judge.json": "b335fcc0ce392b1dd92814822cd323de12ab2aea467f5ae3fcd6d099697a6ada",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/result.json": "2bde961028fa6827ca29b2163778b29095bbfc222abc791f1ae28001d144f26f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/score.json": "8993922ac1c7377965fdd0608a6600a188ce370bd4ac4604e3c4e3e194bfa2f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/metrics.json": "91e3526b43183f5ea3650a438d864384eabc4e9e4de6757a1f1766f2ab7d5c9f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/facts.json": "be94841496d73ff8f57a91eab1a2e41907b397204f341c2d2e4e99e18c60a61a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-35b-a3b-q4-2026-10-21T20-55-40-599Z-31no/codex-inline-judge.json": "50e6cccdc2e3863ac9d63a6ef809a1cc94b04e7073e85af1bd07344db11ccdc9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/result.json": "8ff82dca857abfac08e016735105de6178fc0f117e08741fc349ff40deb74311",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/score.json": "dcde7e185b0ddc2cd94090298bca7a85d117504177ad67a3f5e863bc1ee2b53d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/metrics.json": "12b94bd7d32861432ef4903a81f520a8814946f6c358c2eeb58aeb5942a9cff3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/facts.json": "e5ae73a38531cccb30f40c095bba4d5a6acf6a853ffd3821543f89bd01b2ee03",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-meeting-followup/meeting-followup/meeting-followup-ornith1-5-35b-a3b-q4-2026-10-21T21-00-21-980Z-d3q9/codex-inline-judge.json": "52e9a7ff1991ae134f496139c2768ec31246814f4e752b194c78c6d60075c6ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/result.json": "796c46d0b2ed7ae184e1ffb466a78abc92c4e82c1ac4b339f3c77d05f80d5ef0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/score.json": "397dd03388f95911ca3c48fdfa22c846870b3cc366892e18d45579a79899e668",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/metrics.json": "c7bb47fd53846179625cd9c5778f74483675514fc9d5ccb279eecddc2b10bc80",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/facts.json": "bf4079c843d406904c12ce7d70352ac3aca912ef7796fd55e2785ed99018d294",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-spreadsheet-model/craftbook-spreadsheet-model/craftbook-spreadsheet-model-ornith1-5-35b-a3b-q4-2026-10-21T21-06-34-092Z-ds6k/codex-inline-judge.json": "1c2f839047c4647d465e784c93834b1e03ec21b726ec0bab8f80454a9e5b63c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/result.json": "c8610b083cb59aa87acda96a056aaa01585be9987b2cad758beafb63fbe592f9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/score.json": "538fee7b11b72de691733f468744d50a4694ec321fa7ba682c33c2563f6abdc4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/metrics.json": "76928a028c7dc96429e24785e4e6b4e98fc8343218ffeaff9f37b8312ffa9a9e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/facts.json": "57ccf519cfeacb52769ac0e28cdcddbdc3afe631915ed89341a4dc2101cbe82a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-35b-a3b-q4-2026-10-21T21-32-19-178Z-x998/codex-inline-judge.json": "1f3aadbfb456b8c519629629d51bf50b15a318ab64a343b5ae5d98dff7014224",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/result.json": "66bd2c8395d7709d64794ec3b99d10a29d77b82c49b02289f3fa890e1d8fd770",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/score.json": "1bf6da9d2ed8cdbf6c14830a093b986b3cf71f0781df5f942900501a29907744",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/metrics.json": "bd296e077d9314079f0f48e8db7a1a4c41d3989cd50f22fea81a2dce8bdda8f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/facts.json": "e9d3b35205390dff04800f439486aa2e9262903aa25a7cc669f83696c73a22f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-docblocks-theme-roundtrip/docblocks-theme-roundtrip/docblocks-theme-roundtrip-ornith1-5-35b-a3b-q4-2026-10-21T21-40-00-539Z-r3ld/codex-inline-judge.json": "6d3e41946e9c01496d6b49adc67c6fa2e51259e9902d4bb2f04f736181e7f575",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/result.json": "948137785f13bb2e4eba307469856ef0bd59a7df303b0b15e51c3432947da04a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/score.json": "e74103aa6c569b8a262a3cd8fcc159e451beabee28699a0c17920c2d16869091",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/metrics.json": "f2ef0243fd8e06e3ba2299d97e5436d5c712b240bfc277b6e8cd1b0d7c5a4c19",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/facts.json": "caac9bf9e209fcebf73a3ccbfdb75924dd7bc04d460320e997b6c834ddf086c6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-powerpoint-deck/craftbook-powerpoint-deck/craftbook-powerpoint-deck-ornith1-5-35b-a3b-q4-2026-10-21T21-45-19-227Z-hd3i/codex-inline-judge.json": "1be5c41c37b9b7662de4f14b3fde35285e905cf3698b791ae6afc276fa94af91",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/result.json": "2667c716f9d2478e3f20f6d8178824a410ecac1134e74869ee6023986891e5a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/score.json": "4434cdb17145c87dcfb3ea7d69ebc848a416117014eceec520bf3bcf1c05ffe3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/metrics.json": "cff2ecff94b74e2a3579bc0ee30325ae97a8ab188b95d001c7b4110fd4dbb7ce",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/facts.json": "17c3691b2e215ff3785fbdf56e3a9a5f45c8254264903e0e1eaeaa20bbaa2aa3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-wikipedia-research-brief/wikipedia-research-brief/wikipedia-research-brief-ornith1-5-35b-a3b-q4-2026-10-21T21-53-43-774Z-a5wg/codex-inline-judge.json": "9b33ac599f9a486fb86dc104df7f0d27d794f761f8a73f6cf0a87959e4bbd6fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/result.json": "c0342ac58e569997dbb03bb520b9d396f8f95e72d158501e81b6ad4658a42c10",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/score.json": "977a958c7164e462ecd149f672c48262be0d3ea571e9d355cb70a934b21b9666",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/metrics.json": "2e310f7ec72edef30e1115631059a7229dc4bbe87b2bc9f3c7b669c0d145c571",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/facts.json": "68b7b969221768dd5a10119ec797c94815de6f475904b9468a3b696e037e44ef",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-week-plan/craftbook-week-plan/craftbook-week-plan-ornith1-5-35b-a3b-q4-2026-10-21T21-59-57-653Z-0v5u/codex-inline-judge.json": "146c96bf26786d3669d3c6b93b252d29942899a2042ec5a85c4d3e6c72930dfa",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/result.json": "f8a6265f2b29c395168e753876df7f17ed9ed08a4de07add52d04800e706e24e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/score.json": "bb3148cdbe6de5fa1ea35603db8d6c03d6fd1cd50caf157bbf870fa9c86136a8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/metrics.json": "ae03a9dd6ac6c4593c9dc86b6f9cc61d4d68e02cf2b42962e9ef7579fbceedd9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/facts.json": "91030acfe3b5fc52f7cb6dc3a0e19bf3b35b011e7c21f4f5e4aeb66e17e46984",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-ab-test-readout/craftbook-ab-test-readout/craftbook-ab-test-readout-ornith1-5-35b-a3b-q4-2026-10-21T23-23-35-283Z-l302/codex-inline-judge.json": "58aa855965d2062cddc780af841fae177ab7bea1c5c6b7dc128bcac550940014",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/result.json": "eaa8e627e30ee3b7cfccbed47efb4d6a5553305aa4b6a3def8619a2851d4e91b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/score.json": "cab2fe862a80b7c94b5b32f27556fa5214ec6d5a2b43d3329c6e12c151cf79a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/metrics.json": "c7620fc3907a4f8df3076a4853a485969360b4a60de21abeb415fe33239692b7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/facts.json": "484857a87b33a02fca9850a7659c364b269f22effcbe162cdd9aa035d6782739",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-annotated-bibliography/craftbook-annotated-bibliography/craftbook-annotated-bibliography-ornith1-5-35b-a3b-q4-2026-10-21T23-36-00-226Z-t5os/codex-inline-judge.json": "ba07dacf3701937c3d8f606975ef977d121df905920ee87f9a1a84558ed1c8f1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/result.json": "31b03ba7367badd4557c9aa9fde73da7119513a9926fd55a18a754ef17acd4e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/score.json": "018e7a07c0f4fbc2336e926dbf4c1010929a912543cab2a75bdd90c470485e81",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/metrics.json": "1872a1181b70a70bcec5162dccf0bb2b8f626825ca5032375d3f9b8a1273adc6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/facts.json": "3c26c6b3a71d4cc4881acacf9a7fc70e9c0a996d0837056fbc9b5753e80525c9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-productivity-craftbook-research-to-document/craftbook-research-to-document/craftbook-research-to-document-ornith1-5-35b-a3b-q4-2026-10-21T23-44-17-416Z-j769/codex-inline-judge.json": "17a4bd31e43efe74b7bdc2c1d238ef08c0ce7b396155e9846803d67079acd3c8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/result.json": "587131160816ff834e09e1bbcc25da6152c36fac51168538210196efec60fffd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/score.json": "abc5bd962e9ecfed965a61632d7ac5d0e836b11fe7195d6d19406280f9ef5f0b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/metrics.json": "230e5b49a72491f0d216644debdf529b5eb8a88f837a25259d593152e28f430b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/facts.json": "63a2bf7d370f281dd3c509e90072618fd19e32a557cf5c9c7c20f98c94a8e053",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-fix-squisq-bugs/fix-squisq-bugs/fix-squisq-bugs-ornith1-5-35b-a3b-q4-2026-10-21T23-50-18-410Z-eo0i/codex-inline-judge.json": "0e6d40a6dd93e5d3e34295a320b61f131ec89c33954fc872429a022c1f8b00d2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/result.json": "ea005cf6f08e7322d0ce0a41c0be6584d6d8c389f4d965d63b923e5930c57d0f",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/score.json": "7d6fd51e00c630463a046bb91abb78a4cb31af0ce44146e7c6e7bae3cd6b373c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/metrics.json": "729a0ddcd54899f9432546c1fb4463586546ac1c4cc6f83edad065ae8dff8e4e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/facts.json": "2e46c5532f80133ac3fc50e1e83943a1a7a3db19583335c265c203b7346de2a6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-dev-craftbook-routing/dev-craftbook-routing/dev-craftbook-routing-ornith1-5-35b-a3b-q4-2026-10-21T23-55-34-479Z-ys84/codex-inline-judge.json": "a1791672928060a08212db8ed8253f73df267a60b1be1f4e9718205d785f1bf3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/result.json": "525c9d718fc03d3238c8c9cb2590fb9ecf941c801fa5f3c4e0f457c80e1e097c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/score.json": "40a57e23b208466681895f6256380eea02763afd7211d0b0b354dfe3bbe47434",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/metrics.json": "5aaee590cd46f3e93b4299a813516bc40dfe372bdc51fc75392b39b1ce0af7ad",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/facts.json": "f9302a0bcfacc05b0d30415b605d29ff84b80354b21b9e1569982d9c49c99779",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-interface-contract/interface-contract/interface-contract-ornith1-5-35b-a3b-q4-2026-10-22T00-01-12-440Z-toj0/codex-inline-judge.json": "c462014059555e313d3f0bd9f12cafb1bf9b5ed264771d1a507c3bffff888dcb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/result.json": "e433b141c2abfc4960f7ac951bbc28e43ca44da91ee2f10d4d070b165b9435fd",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/score.json": "354ea8c1a9ac70877631cb4c0290f2d569fd9612841f605500bd8440304a08d8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/metrics.json": "a4d4acec1d2a98bfc1471aea17e6d76270bc5d673bffe922c7fd972cb02d6715",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/facts.json": "925da44b81527662906bdbc505294c84f0c0c228d661d36f2220f83f6502cd05",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-code-review/craftbook-code-review/craftbook-code-review-ornith1-5-35b-a3b-q4-2026-10-22T00-07-59-545Z-ajjt/codex-inline-judge.json": "abe8c86c24c03f3e89f15d013c40c0c8acccb4443637ca68969ee366c74f3449",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/result.json": "39bef1230b6fc17c41db39eea3b30aee625b314e94d2060820dc848548adb806",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/score.json": "7f8b07084bc911560db31a060942e50d42c346bc8d99c26997dbc5237b54b149",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/metrics.json": "11124b5e26d9daf3532a634d652ef827ed899a125740ae22bdaf9b96d7645fae",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/facts.json": "20833956911b6d9f093db747a2fb88a2079655e2ec6b9447af48b4b2dfa111c0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-api-contract-review/craftbook-api-contract-review/craftbook-api-contract-review-ornith1-5-35b-a3b-q4-2026-10-22T00-15-58-715Z-9a1z/codex-inline-judge.json": "2f33ced646e7f8242bb184f24bfdc37ede6d3a1f33fc2523e836b2381f924795",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/result.json": "c1bff553cd415cf135a7421c5293200cc03e209497964d72a490ef09c77a7bab",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/score.json": "f891b4032c62fef5f8d09e15018e22b505525d75f710845fd2222e78e482de8d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/metrics.json": "8be6bd5a6da0a31560a715ef033e8b6b9300c10565202a27014e52969094bbcb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/facts.json": "567dfefe1cc37b6a21d904502980e6d2f2339892f963667951adbac88b894797",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-deep-security-review/craftbook-deep-security-review/craftbook-deep-security-review-ornith1-5-35b-a3b-q4-2026-10-22T00-29-22-679Z-e65k/codex-inline-judge.json": "e00e5c9fbc1cde0187df4048a8ff11afa21912044cf6dc02eb262378e8af1929",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/result.json": "cbc27d9eb192d5f3b96a3bc1e33fac8004b3fb1a6e39d046a93f5d7a39a29be0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/score.json": "cb18e187ab7adc4765e75e09ea51543251393bd0ae197975c1788e209dfe8f76",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/metrics.json": "3dea55a47b8a063d472003de0f56252692174769244de244b5953c5d819aec18",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/facts.json": "eef94bff67b2fd842574bbae194341ed56b172c2b24c13b4746af691979e31d4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-large-pr-review/large-pr-review/large-pr-review-ornith1-5-35b-a3b-q4-2026-10-22T00-36-53-350Z-gzs8/codex-inline-judge.json": "9f0103c0ad1fced04fe3cb652a8efa16b6ec03f3c160afd9fe8cefd2a6954830",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/result.json": "0f99cc996c1a70d7799cdc08c15d4819cac1769601a6f627b3dd56fa1c659b40",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/score.json": "761b5731f08e702dd6d63ccd2fb725a1a295d052448edd5d6f131e7cd152b730",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/metrics.json": "57852e27e3dd1eb6c6b35150a698484829d74019aa45a80783e4dbfc93522d6b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/facts.json": "027dcaee19f070419dc011697ac3de14c8ba995ac37cfcaf2bbfd29372faaf1d",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-codemod-sweep/craftbook-codemod-sweep/craftbook-codemod-sweep-ornith1-5-35b-a3b-q4-2026-10-22T00-57-34-971Z-coht/codex-inline-judge.json": "be1f230738b512a8943d2aaaf1f446654773725267b62b6842ba32683472d6cf",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/result.json": "d1b5c98deeacf9e005cf815e17b458ae15c967ef89e39eab62aa6a85834e13a7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/score.json": "c055e545e0ea64b2659f968e77644ed57b1d3bb6a6b38cc75136ad38b3323b8b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/metrics.json": "c1dd4a9f4150fe9f1953fc67e8f0cb7b5b3778ee86ca85b933ebb4b7894e667a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/facts.json": "b4c7f9e95fe6094bdad948d12eb8975fa6951c944cc8f78d9e0c5eaaf7d4d8d8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-refactor-module/craftbook-refactor-module/craftbook-refactor-module-ornith1-5-35b-a3b-q4-2026-10-22T01-05-23-877Z-poml/codex-inline-judge.json": "af525c0dd64fc323a692a2a8b6afec4940dcc45585771f0bf554cc7ca908cdd2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/result.json": "26f51d893dadeed2c06927b0d24635ee5762aa92157b039c2b96fdc91f42d7f5",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/score.json": "438b0fc7dfbc172947abaa320f545fe74b28c9c3e4994a306226e5870259c861",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/metrics.json": "e1bc52600c464c785181288a92e5ee3fd1237b858c82d589af9df0af5ddf0f9b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/facts.json": "1936366e18f79c0297473e8e7e33182e5505bf31ff7430d5a6a44fcfd82bdae6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-developer-craftbook-bug-fix-tdd/craftbook-bug-fix-tdd/craftbook-bug-fix-tdd-ornith1-5-35b-a3b-q4-2026-10-22T01-22-39-862Z-1nfx/codex-inline-judge.json": "023e74da5a27fd3c5d2c0392ee07384cab11112d75daa8d14ac531dd56e7b7e0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/result.json": "eec8bd97357819a14ed0865d7ce3abd74a9e3f60e9a573d8a169de61f860bb95",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/score.json": "79399128c2a83e4c36b8e1a001383f3836b47174d073b4b4a54080d23d1fb4f0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/metrics.json": "a0a6e0db0ea88aa5fa29f524d3144cc4c393a8b8bf36cf316b93911daee2c643",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/facts.json": "18008dcec5d32c97f45b49dfedb915a5f509b72f1de20ac15dd24eb24a932f30",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-find-vs-create/craftbook-find-vs-create/craftbook-find-vs-create-ornith1-5-35b-a3b-q4-2026-10-22T01-35-18-297Z-qrxs/codex-inline-judge.json": "dfb388faf937db7390fd3dd62e63f939ddff2c25de57d0ec215959c87544d58a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/result.json": "b6a16130f6dd9c1a3be24baf41c1c9c5d09809338562d4edbd631e89034d6142",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/score.json": "1d76f74573dce7f8f0252aade87abcf2e713850d8d5d7b41091033522f3bd38a",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/metrics.json": "5c24b32d90a74f4df24f5d4f8ffc1582747ec28a78d0a2ecd3b02d6e4b80692b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/facts.json": "714e1779f9860d128d8837d83bfc090123ac0262cf74d6f5dd8425acfc55995c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-edit-midtask/craftbook-edit-midtask/craftbook-edit-midtask-ornith1-5-35b-a3b-q4-2026-10-22T01-39-53-765Z-iu6p/codex-inline-judge.json": "628a04d6b1ef4a975fcbd075360715a5e99dc82c5206617e20479002e32ec4f3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/result.json": "2e15f672500a37eb886f97131e9c38327603a9afd3da54fb343c18293990dbda",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/score.json": "f06dff21029719175a632a423df3195dd1e26412d082e1d3f2004c90197a3dba",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/metrics.json": "63ab04e06f804631e69cc396a4cef37f2447c15962baf8d0e992b434afef81b4",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/facts.json": "2222b5c1bdc2d424465e28a4e53a94b27fc0f964b350067aaed7dcda9c701ec9",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-invoice-run/craftbook-invoice-run/craftbook-invoice-run-ornith1-5-35b-a3b-q4-2026-10-22T01-46-47-348Z-v11v/codex-inline-judge.json": "b1f9ecf86890cdc291d7ee7b6c77c7be29146ad1ed9f731f8048260e666aeea3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/result.json": "49443a61228b4da241b4424c37024aada6149e93e7584eac761829b4ae456fa2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/score.json": "f0ec9106181ac3718fb9e9fd9d257f24e1f99dad6ab00e83b04df1e995387a50",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/metrics.json": "085065e685a7b5ec6d67ab1f3d20e99a17c757bfe2753f0d053d6aaccd65aa5e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/facts.json": "6eb9800096de13e8e96409dce4a3fa44a4c9b8425dca7f89e8d3ae56ff35602c",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-linear/craftbook-author-linear/craftbook-author-linear-ornith1-5-35b-a3b-q4-2026-10-22T01-54-13-868Z-aw8i/codex-inline-judge.json": "700c03ce9c19aaf6f6d8031a82a1942e7b7f1f72d60a34c383dc6cd30297d878",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/result.json": "bc7d7bae2ee592053b0ea17436535c49d6e551d68e66d43114f811beda260fc7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/score.json": "d05e41232687a9aa90898227fbba365015211ad6d1f61f7d68a7efec14a42893",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/metrics.json": "be42d4eb798a5ed71544b88ae1874df58537a3a9a73b999a8c20ebacd97a1288",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/facts.json": "5fe3f1b690209fd087c9b9f4f11aa68b2f26c63117204f65bb30a7390118b8e2",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-route-multi/craftbook-route-multi/craftbook-route-multi-ornith1-5-35b-a3b-q4-2026-10-22T02-43-02-868Z-0u9q/codex-inline-judge.json": "7cea0cb8e534447f974e5ad66736eea198dda0ae8db8a8a8b741e6f31d3373c3",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/result.json": "c98d7358fd46b43ed5fa5863270402dfce020251bb20e0e1c2a19d64565d8636",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/score.json": "1e0d9e8712c7cc2a31034e6856956b4ae5d8ef50c418b6a9c88f92ce7f343890",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/metrics.json": "66e629fb789df0a406526eee4c4e60e96f30cac1806b5574dc1b8bc469f32045",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/facts.json": "71acb722a4a55666ddae457d423d5d2c037af3f6687c65feee6421a7f71610e1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-export-generalize/craftbook-export-generalize/craftbook-export-generalize-ornith1-5-35b-a3b-q4-2026-10-22T02-50-20-192Z-cxkn/codex-inline-judge.json": "6eb0979f35dfd5ebb6187b3c45e15a2069584d5de25408b912fcd8453414e1eb",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/result.json": "50724c3ac56306fe7af436bd9184a042ae5df61f3ec466fd1aa3bf4362b4b85b",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/score.json": "ee320654e8a49eeeaf8678593118b05f624946462e4cd0323a7a3c54b28741f6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/metrics.json": "e3d68c5c0fa9644d8696a1a7188862f7dd1de278cc21e262e5b057faa7d24fe0",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/facts.json": "e065198506ca4593f500bde50fac78e83900703fe27914f7641f1a3f3e9868f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-params/craftbook-author-params/craftbook-author-params-ornith1-5-35b-a3b-q4-2026-10-22T02-59-35-070Z-1phy/codex-inline-judge.json": "ff1d492efe6630655ffa3f606aeab596e54f3d4b0bb644efd9da5237fe8af1c1",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/result.json": "75744c7ab8444a86f2c327d977caba1596cfc9e36bbaa2e1afd9b0c977cbe9f8",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/score.json": "d4771512780113809d95a75ea29e278aa3aab3438e385fde0f101bd6927bcb4e",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/metrics.json": "fe1ff217eda3c45a7b7f82ef07481de868efde26ea8f708c95d5713dc97867c7",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/facts.json": "556fc6f46f06c0a4c3097ca8aafa5deba1df7990716b3febb398bd13951897b6",
      "evals/runs/scorecard-local-seven-baseline-2026-10-15/ornith1.5-35b-a3b-q4/r2-ornith1.5-35b-a3b-q4-complex-work-craftbook-author-gate-script/craftbook-author-gate-script/craftbook-author-gate-script-ornith1-5-35b-a3b-q4-2026-10-22T04-20-00-701Z-17st/codex-inline-judge.json": "c34bd960214a0f6bd7f1181a989adc04c6d7b38520e704fee22068e3d406e324",
      "evals/runs/scorecard-seven-models-2026-10-15-proposal/memory-admission-observation-2026-10-22T0257.json": "a282f767c955678c5e9c884642fcbe4b116bef8a6404abaccb15e5738f2f18a1"
    }
  }
}